Dataset Viewer
Auto-converted to Parquet Duplicate
year
int32
1.99k
1.99k
subset
stringclasses
4 values
file
stringclasses
14 values
variable
stringclasses
202 values
code
stringclasses
96 values
code_end
stringclasses
0 values
description
stringclasses
408 values
codebook_description
stringclasses
166 values
published_unweighted_n
float64
5
12.1k
โŒ€
published_weighted_pct
float64
n_in_file
float64
0
12.3k
โŒ€
published_n_matches
bool
1 class
value_source
stringclasses
3 values
1,991
SADC_HS_District
sadc_2023_district
sitecode
AB
null
Albuquerque, NM
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
AK
null
Alaska
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
AL
null
Alabama
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
AR
null
Arkansas
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
AZB
null
Arizona
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
CA
null
California
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
CC
null
Clark County, NV
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
CE
null
Cleveland, OH
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
CH
null
Chicago, IL
null
null
null
1,529
null
sas_format;data
1,991
SADC_HS_District
sadc_2023_district
sitecode
CM
null
Charlotte-Mecklenburg County, NC
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
CO
null
Colorado
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
CT
null
Connecticut
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
DA
null
Dallas, TX
null
null
null
3,331
null
sas_format;data
1,991
SADC_HS_District
sadc_2023_district
sitecode
DE
null
Delaware
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
DU
null
Duval County, FL
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
EA
null
Eaton Consortium, MI
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
FL
null
Florida
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
FT
null
Broward County, FL
null
null
null
1,307
null
sas_format;data
1,991
SADC_HS_District
sadc_2023_district
sitecode
FW
null
Fort Worth, TX
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
GA
null
Georgia
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
GE
null
Genesee Consortium, MI
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
GS
null
Gaston County, NC
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
HI
null
Hawaii
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
HL
null
Hillsborough County, FL
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
HO
null
Houston, TX
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
IA
null
Iowa
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
ID
null
Idaho
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
IL
null
Illinois
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
IN
null
Indiana
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
KS
null
Kansas
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
KY
null
Kentucky
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
LA
null
Louisiana
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
LO
null
Los Angeles, CA
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
MA
null
Massachusetts
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
MD
null
Maryland
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
ME
null
Maine
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
MI
null
Michigan
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
ML
null
Milwaukee, WI
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
MM
null
Miami-Dade County, FL
null
null
null
2,150
null
sas_format;data
1,991
SADC_HS_District
sadc_2023_district
sitecode
MO
null
Missouri
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
MS
null
Mississippi
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
MT
null
Montana
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
NC
null
North Carolina
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
ND
null
North Dakota
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
NE
null
Nebraska
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
NH
null
New Hampshire
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
NJ
null
New Jersey
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
NM
null
New Mexico
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
NO
null
New Orleans, LA
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
NV
null
Nevada
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
NW
null
Newark, NJ
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
NY
null
New York
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
NYA
null
New York (Excluding NYC)
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
NYC
null
New York City, NY
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
NYG
null
Borough of Bronx, NY
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
NYH
null
Borough of Brooklyn, NY
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
NYI
null
Borough of Manhattan, NY
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
NYJ
null
Borough of Queens, NY
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
NYK
null
Borough of Staten Island, NY
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
OA
null
Oakland, CA
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
OK
null
Oklahoma
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
OL
null
Orange County, FL
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
PA
null
Pennsylvania
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
PB
null
Palm Beach County, FL
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
PH
null
Philadelphia, PA
null
null
null
1,429
null
sas_format;data
1,991
SADC_HS_District
sadc_2023_district
sitecode
PO
null
Portland, OR
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
PS
null
Pasco County, FL
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
RI
null
Rhode Island
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
SA
null
San Diego, CA
null
null
null
658
null
sas_format;data
1,991
SADC_HS_District
sadc_2023_district
sitecode
SB
null
San Bernardino, CA
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
SC
null
South Carolina
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
SD
null
South Dakota
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
SE
null
Seattle, WA
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
SF
null
San Francisco, CA
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
SP
null
Spartanburg County, SC
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
ST
null
Shelby County, TN
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
TN
null
Tennessee
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
TX
null
Texas
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
UT
null
Utah
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
VA
null
Virginia
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
VT
null
Vermont
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
WI
null
Wisconsin
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
WV
null
West Virginia
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
WY
null
Wyoming
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sitecode
XX
null
United States
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
age
1
null
12 years old or younger
null
null
null
12
null
sas_format;data
1,991
SADC_HS_District
sadc_2023_district
age
2
null
13 years old
null
null
null
25
null
sas_format;data
1,991
SADC_HS_District
sadc_2023_district
age
3
null
14 years old
null
null
null
1,052
null
sas_format;data
1,991
SADC_HS_District
sadc_2023_district
age
4
null
15 years old
null
null
null
2,537
null
sas_format;data
1,991
SADC_HS_District
sadc_2023_district
age
5
null
16 years old
null
null
null
2,969
null
sas_format;data
1,991
SADC_HS_District
sadc_2023_district
age
6
null
17 years old
null
null
null
2,319
null
sas_format;data
1,991
SADC_HS_District
sadc_2023_district
age
7
null
18 years old or older
null
null
null
1,479
null
sas_format;data
1,991
SADC_HS_District
sadc_2023_district
age
other
null
** Data Error **
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
age
.
null
Missing
null
null
null
11
null
sas_format;data
1,991
SADC_HS_District
sadc_2023_district
sex
1
null
Female
null
null
null
5,413
null
sas_format;data
1,991
SADC_HS_District
sadc_2023_district
sex
2
null
Male
null
null
null
4,970
null
sas_format;data
1,991
SADC_HS_District
sadc_2023_district
sex
other
null
** Data Error **
null
null
null
0
null
sas_format
1,991
SADC_HS_District
sadc_2023_district
sex
.
null
Missing
null
null
null
21
null
sas_format;data
1,991
SADC_HS_District
sadc_2023_district
grade
1
null
9th
null
null
null
3,022
null
sas_format;data
1,991
SADC_HS_District
sadc_2023_district
grade
2
null
10th
null
null
null
3,310
null
sas_format;data
End of preview. Expand in Data Studio

YRBSS 1991โ€“2025 in Parquet (with dictionaries and survey-design metadata)

This dataset contains the public data files of the U.S. Youth Risk Behavior Surveillance System (YRBSS) of the Centers for Disease Control and Prevention (CDC): every national high school Youth Risk Behavior Survey (YRBS) from 1991 to 2025, the 1995 National College Health Risk Behavior Survey, and the CDC combined datasets (SADC) of state and local (district) high school and middle school surveys. Each fixed-width ASCII file has been read with CDC's own SAS input program and converted to Parquet under its original CDC file name. It also includes:

  • a variable dictionary for each survey year, built from the CDC SAS input and format programs and the codebook of the national data user's guides. It gives the data location, SAS label, question text, value labels and the frequency and weighted percentage CDC published for each code. For the dichotomous QN variables it adds the numerator and denominator (the analytic universe);
  • metadata/FILES.parquet (every file: rows, grain, join keys, sites, source URLs, MD5) and metadata/YEARS.parquet (every survey year: weight, stratum and PSU variables, design and pooling guidance).

Built for the Bioanalysis R package. It works equally well with any tool that reads Parquet: DuckDB, Arrow, Polars, pandas or R.

Values are the original public-use codes; nothing is recoded, imputed or dropped. See Data by category below for what is included, Hugging Face subsets and splits for the viewer tables and Data dictionary for variable and value descriptions.

Source: CDC. This compilation is not endorsed by CDC, HHS or the U.S. Government. The original files are available at no charge from the CDC YRBSS data page.

Layout

data/<year>/<subset>/<FILE>.parquet
data/<year>/Dictionary/VARIABLES.parquet   one row per file x variable
data/<year>/Dictionary/VALUES.parquet      one row per file x variable x code
metadata/FILES.parquet                     every file: rows, grain, join keys, sites, URLs, MD5
metadata/YEARS.parquet                     every year x subset x release: design variables
pipeline/                                  the R scripts that rebuilt everything from CDC

<year> is the survey year; <subset> (folder) and <FILE> keep CDC's names. Each Combined Datasets (SADC) file is split by its year variable, so data/*/SADC_HS_State/sadc_2025_state_*.parquet is the complete CDC file. A site that withdrew is not in the later release, so each Combined Datasets release is kept (see below).

Data by category

The data are grouped as CDC's YRBSS is organised: the national survey, state and district (local) surveys of high school students, and middle school surveys. A file type is a CDC file name with its year replaced by <YYYY>. National files are named after the survey year: yrbs<YYYY> and XXH<YYYY>_YRBS_Data (1991โ€“2015 and 2017โ€“2025 respectively). Combined Datasets (SADC) files are named after their release year (sadc_2025_* is the 2025 release, sadc_ms_* is middle school) and contain every survey year from 1991 (high school) or 1995 (middle school) to the release; here each is split by its year variable into one Parquet file per survey year, so data/<year>/<subset>/<FILE>.parquet keeps the CDC file name. State files carry CDC's alphabetical suffix (_a_d = states Aโ€“D by state name).

Category Contents File types Parquet files Records Survey years
National CDC's national school-based survey of students in grades 9โ€“12 (one file per survey year), the 1995 National College Health Risk Behavior Survey, and the national surveys again inside the Combined Datasets (standardized variables, all years in one file). 3 54 790,259 1991โ€“2025
State State high school surveys conducted by state health and education departments, from the CDC Combined Datasets; CDC splits them into files by state name. 7 232 2,765,184 1991โ€“2025
District (local) Surveys of large urban school districts (high school), from the CDC Combined Datasets. 1 35 835,197 1991โ€“2025
Middle school State and district middle school surveys from the CDC Combined Datasets (there is no national middle school survey). 3 41 468,057 1995โ€“2023

Records count each Combined Datasets release separately (the national surveys appear in the national files and again in sadc_*_national). Expand a category to see every file type with its subset (folder), its title, the survey years it covers, its records and its key.

National: 3 file types, 54 Parquet files
File type Subset (folder) Title (most recent release) CDC files Survey years Records (all years; SADC: latest release) Key
nchrbs<YYYY> National_College_YRBS National College Health Risk Behavior Survey 1995 nchrbs1995 1995 4,814 source_row (within this file only)
yrbs<YYYY>, XXH<YYYY>_YRBS_Data National_YRBS National Youth Risk Behavior Survey 2025 (high school) yrbs1991 โ€ฆ XXH2025_YRBS_Data 1991โ€“2025 265,385 source_row (within this file only)
sadc_<YYYY>_national SADC_HS_National YRBS Combined Datasets, 2025 release (high school): national YRBS, survey years 1991-2025 sadc_2023_national, sadc_2025_national 1991โ€“2025 (releases 2023, 2025) 520,060 sitecode + year + record (also: source_row within the CDC file)
State: 7 file types, 232 Parquet files
File type Subset (folder) Title (most recent release) CDC files Survey years Records (all years; SADC: latest release) Key
sadc_<YYYY>_state_a_d SADC_HS_State YRBS Combined Datasets, 2025 release (high school): state surveys, states A-D (by state name) sadc_2023_state_a_d, sadc_2025_state_a_d 1991โ€“2025 (releases 2023, 2025) 167,409 sitecode + year + record (also: source_row within the CDC file)
sadc_<YYYY>_state_e_h SADC_HS_State YRBS Combined Datasets, 2025 release (high school): state surveys, states E-H (by state name) sadc_2023_state_e_h, sadc_2025_state_e_h 1991โ€“2025 (releases 2023, 2025) 177,185 sitecode + year + record (also: source_row within the CDC file)
sadc_<YYYY>_state_i_l SADC_HS_State YRBS Combined Datasets, 2025 release (high school): state surveys, states I-L (by state name) sadc_2023_state_i_l, sadc_2025_state_i_l 1991โ€“1997, 2001โ€“2025 (releases 2023, 2025) 236,389 sitecode + year + record (also: source_row within the CDC file)
sadc_<YYYY>_state_m SADC_HS_State YRBS Combined Datasets, 2025 release (high school): state surveys, states whose names start with M sadc_2023_state_m, sadc_2025_state_m 1993โ€“2025 (releases 2023, 2025) 691,221 sitecode + year + record (also: source_row within the CDC file)
sadc_<YYYY>_state_n_p SADC_HS_State YRBS Combined Datasets, 2025 release (high school): state surveys, states N-P (by state name) sadc_2023_state_n_p, sadc_2025_state_n_p 1991โ€“2025 (releases 2023, 2025) 723,026 sitecode + year + record (also: source_row within the CDC file)
sadc_<YYYY>_state_q_t SADC_HS_State YRBS Combined Datasets, 2025 release (high school): state surveys, states Q-T (by state name) sadc_2023_state_q_t, sadc_2025_state_q_t 1991โ€“2025 (releases 2023, 2025) 190,445 sitecode + year + record (also: source_row within the CDC file)
sadc_<YYYY>_state_u_z SADC_HS_State YRBS Combined Datasets, 2025 release (high school): state surveys, states U-Z (by state name) sadc_2023_state_u_z, sadc_2025_state_u_z 1991โ€“2025 (releases 2023, 2025) 579,509 sitecode + year + record (also: source_row within the CDC file)
District (local): 1 file type, 35 Parquet files
File type Subset (folder) Title (most recent release) CDC files Survey years Records (all years; SADC: latest release) Key
sadc_<YYYY>_district SADC_HS_District YRBS Combined Datasets, 2025 release (high school): district (large urban school district) surveys sadc_2023_district, sadc_2025_district 1991โ€“2025 (releases 2023, 2025) 835,197 sitecode + year + record (also: source_row within the CDC file)
Middle school: 3 file types, 41 Parquet files
File type Subset (folder) Title (most recent release) CDC files Survey years Records (all years; SADC: latest release) Key
sadc_ms_<YYYY>_district SADC_MS_District YRBS Combined Datasets, 2023 release (middle school): district (large urban school district) surveys sadc_ms_2023_district 1995, 1999โ€“2023 (release 2023) 141,429 sitecode + year + record (also: source_row within the CDC file)
sadc_ms_<YYYY>_state_a_m SADC_MS_State YRBS Combined Datasets, 2023 release (middle school): state surveys, states A-M (by state name) sadc_ms_2023_state_a_m 1997โ€“2003, 2007โ€“2023 (release 2023) 112,714 sitecode + year + record (also: source_row within the CDC file)
sadc_ms_<YYYY>_state_n_z SADC_MS_State YRBS Combined Datasets, 2023 release (middle school): state surveys, states N-Z (by state name) sadc_ms_2023_state_n_z 1995, 1999โ€“2023 (release 2023) 213,914 sitecode + year + record (also: source_row within the CDC file)

Two Combined Datasets releases (2023 and 2025). CDC includes a state or district only with its permission, and the 2025 release leaves out sites of the 2023 release:

  • SADC_HS_State: 157 site-years (629,479 records) only in the 2023 release: AK AL AR CO FL GA IA ID MD NC NJ NYA PA TX WY.
  • SADC_HS_District: 129 site-years (231,276 records) only in the 2023 release: CC CM DA DU FT FW HL ML MM NO OA OL PB PH PS SB SP.

Use one release at a time; CDC record IDs differ between releases.

Hugging Face subsets and splits

On Hugging Face a subset (also called a config) is one named table: it is the drop-down list above the Dataset Viewer and the second argument of load_dataset("jamesjiadazhan/YRBSS", "<subset>"). A split is a part of a subset, chosen with split=. In this dataset a split is either all (the whole table) or one survey year (year_2019 holds survey year 2019). Every row also carries its survey year (year in the Combined Datasets and the dictionaries; the national files are one year each), so splits can be stacked.

Subset What it holds Splits
files Catalogue of every data file: survey year, subset, CDC file, release, title, rows, key, sites, path, source URLs, MD5. all
years One row per survey year ร— subset ร— release: rows, files, sites, weight / stratum / PSU variable names, design and pooling guidance. all
dictionary_variables Variable dictionary: one row per file ร— variable. 18, one per survey year (year_1991 โ€ฆ year_2025)
dictionary_values Value dictionary: one row per file ร— variable ร— code. 18, one per survey year (year_1991 โ€ฆ year_2025)
national_yrbs_1991 National Youth Risk Behavior Survey 1991 (high school) (yrbs1991, 12,272 records). all
national_yrbs_1993 National Youth Risk Behavior Survey 1993 (high school) (yrbs1993, 16,296 records). all
national_yrbs_1995 National Youth Risk Behavior Survey 1995 (high school) (yrbs1995, 10,904 records). all
national_yrbs_1997 National Youth Risk Behavior Survey 1997 (high school) (yrbs1997, 16,262 records). all
national_yrbs_1999 National Youth Risk Behavior Survey 1999 (high school) (yrbs1999, 15,349 records). all
national_yrbs_2001 National Youth Risk Behavior Survey 2001 (high school) (yrbs2001, 13,601 records). all
national_yrbs_2003 National Youth Risk Behavior Survey 2003 (high school) (yrbs2003, 15,214 records). all
national_yrbs_2005 National Youth Risk Behavior Survey 2005 (high school) (yrbs2005, 13,917 records). all
national_yrbs_2007 National Youth Risk Behavior Survey 2007 (high school) (yrbs2007, 14,041 records). all
national_yrbs_2009 National Youth Risk Behavior Survey 2009 (high school) (yrbs2009, 16,410 records). all
national_yrbs_2011 National Youth Risk Behavior Survey 2011 (high school) (yrbs2011, 15,425 records). all
national_yrbs_2013 National Youth Risk Behavior Survey 2013 (high school) (yrbs2013, 13,583 records). all
national_yrbs_2015 National Youth Risk Behavior Survey 2015 (high school) (yrbs2015, 15,624 records). all
national_yrbs_2017 National Youth Risk Behavior Survey 2017 (high school) (XXH2017_YRBS_Data, 14,765 records). all
national_yrbs_2019 National Youth Risk Behavior Survey 2019 (high school) (XXH2019_YRBS_Data, 13,677 records). all
national_yrbs_2021 National Youth Risk Behavior Survey 2021 (high school) (XXH2021_YRBS_Data, 17,232 records). all
national_yrbs_2023 National Youth Risk Behavior Survey 2023 (high school) (XXH2023_YRBS_Data, 20,103 records). all
national_yrbs_2025 National Youth Risk Behavior Survey 2025 (high school) (XXH2025_YRBS_Data, 10,710 records). all
national_college_1995 National College Health Risk Behavior Survey 1995 (nchrbs1995, 4,814 records). all
sadc_2025_national YRBS Combined Datasets, 2025 release (high school): national YRBS (sadc_2025_national, 265,385 records). 18, one per survey year (year_1991 โ€ฆ year_2025)
sadc_2025_state YRBS Combined Datasets, 2025 release (high school): state surveys, all state files (sadc_2025_state_a_d, sadc_2025_state_e_h, sadc_2025_state_i_l, sadc_2025_state_m, sadc_2025_state_n_p, sadc_2025_state_q_t, sadc_2025_state_u_z, 1,211,275 records). 18, one per survey year (year_1991 โ€ฆ year_2025)
sadc_2025_district YRBS Combined Datasets, 2025 release (high school): district (large urban school district) surveys (sadc_2025_district, 312,869 records). 18, one per survey year (year_1991 โ€ฆ year_2025)
sadc_2023_national YRBS Combined Datasets, 2023 release (high school): national YRBS (sadc_2023_national, 254,675 records). 17, one per survey year (year_1991 โ€ฆ year_2023)
sadc_2023_state YRBS Combined Datasets, 2023 release (high school): state surveys, all state files (sadc_2023_state_a_d, sadc_2023_state_e_h, sadc_2023_state_i_l, sadc_2023_state_m, sadc_2023_state_n_p, sadc_2023_state_q_t, sadc_2023_state_u_z, 1,553,909 records). 17, one per survey year (year_1991 โ€ฆ year_2023)
sadc_2023_district YRBS Combined Datasets, 2023 release (high school): district (large urban school district) surveys (sadc_2023_district, 522,328 records). 17, one per survey year (year_1991 โ€ฆ year_2023)
sadc_ms_2023_state YRBS Combined Datasets, 2023 release (middle school): state surveys, all state files (sadc_ms_2023_state_a_m, sadc_ms_2023_state_n_z, 326,628 records). 15, one per survey year (year_1995 โ€ฆ year_2023)
sadc_ms_2023_district YRBS Combined Datasets, 2023 release (middle school): district (large urban school district) surveys (sadc_ms_2023_district, 141,429 records). 14, one per survey year (year_1995 โ€ฆ year_2023)

The national files are one subset per year (national_yrbs_<YYYY>) because their questions and columns change from year to year; the Combined Datasets keep the same columns in every year, so each CDC file (or set of state files) is one subset split by survey year.

All 362 Parquet data files belong to a subset. Any file can also be read by path, for example hf://datasets/jamesjiadazhan/YRBSS/data/2019/SADC_HS_State/sadc_2025_state_m.parquet (see Quick start); the Files and versions tab shows every file.

Data dictionary

Each survey year has its own dictionary (1991โ€“2025), covering every file of that year (the national file and the year's part of each Combined Datasets file). It is built from CDC's SAS input and format programs and, for the national files, the codebook and appendices of the data user's guide (PDF):

File Subset One row per Main columns
data/<year>/Dictionary/VARIABLES.parquet dictionary_variables file ร— variable file, file_label (file title), variable, label (SAS label), question_text, universe (QN denominator: who is counted), numerator, summary_text, data_start / data_end (columns in the CDC .dat), type, value_format, n_nonmissing, dictionary_source, sas_input_url, guide_url
data/<year>/Dictionary/VALUES.parquet dictionary_values file ร— variable ร— code code (as a string; null = missing), description (SAS value label, else codebook), published_unweighted_n, published_weighted_pct (CDC codebook), n_in_file (count in this file), published_n_matches, value_source
Source Files Gives
SAS input program national/college 1991โ€“2025; Combined Datasets (all years) location, informat, type, label
SAS format program national/college 1991โ€“2025; Combined Datasets (all years) value labels
Guide codebook (PDF appendix) national/college 1991โ€“2023 question text, published unweighted n (and weighted %) per code
Guide question list (Appendix A), where the codebook has no question text national/college 2017โ€“2021, 2025 question text
Guide QN definitions national/college 2005โ€“2025 numerator, denominator (universe), summary text

Question text and published frequencies are read from PDF text: wrapped text of a variable without a value table (e.g. height, weight) can spill into its neighbour, and published_n_matches = false flags codebook rows misread from the PDF.

Example: file = 'XXH2023_YRBS_Data', variable = 'QN26' (Felt sad or hopeless) has universe Students who answered A or B for Q26; its codes are in VALUES.parquet of 2023. Join a dictionary to data on file + variable. The official CDC documents of each file (data user's guide, SAS programs) are linked in guide_url and sas_input_url here and in the files subset; the CDC listing is https://www.cdc.gov/yrbs/data/index.html.

CDC's SAS programs leave the label of 5 variables blank in 9 files (q63, q62, q61, sexid, sexid2 in sadc_2025_district, sadc_2025_national, sadc_2025_state_a_d, sadc_2025_state_e_h, sadc_2025_state_i_l, sadc_2025_state_m, sadc_2025_state_n_p, sadc_2025_state_q_t, sadc_2025_state_u_z); their labels stay missing and are marked in dictionary_source. They are not filled from another release because question numbers differ between releases.

Values

Values are exactly those of CDC's SAS input programs:

  • Character variables ($ informats, e.g. Q1โ€“Q99, RACEETH) are strings. Leading and trailing blanks are removed, and blank fields are null (SAS missing).
  • Numeric variables (WEIGHT, STRATUM, PSU, QN* in recent years, all SADC numeric variables) are doubles. Blank or . fields are null. Implied decimals of w.d informats are applied only when a field has no decimal point (none occur in these files).
  • Response codes are kept: "1", "2", โ€ฆ for A, B, โ€ฆ in the national files. Numeric codes are kept in SADC. Look them up in VALUES.parquet.
  • In some years (e.g. 2003โ€“2005) the character QN variables use "." for missing. It is kept as published, and CDC's codebook treats it as missing.
  • Variable-name case follows each CDC program (WEIGHT in most years, Weight/Stratum in 2003โ€“2007; lower case in SADC). SAS names are case-insensitive.
  • One column is added to every file: source_row, the 1-based line number of the record in the original CDC .dat file.
  • Two 1990s files (yrbs1997.dat, nchrbs1995.dat) end with a DOS end-of-file character (Ctrl-Z) on a line of its own. That line is not a record and is not included; the CDC record counts exclude it.

Keys and joining

YRBS is cross-sectional. There is no person identifier across years, files or releases, and records cannot be linked between surveys.

  • National and college files have no record ID. The key is source_row, which is unique within each file. It identifies a record only within that file.
  • SADC files: record (CDC record ID) is unique within each file and the key is sitecode + year + record. source_row also identifies the line in the CDC file. The SADC national file repeats the national YRBS records in a standardized layout, but it shares no ID with the national files and its row order differs. Use one or the other.
  • Dictionary: join VARIABLES/VALUES to data on file + variable, and to values on code (as a string). For SADC numeric variables, format the number as text, e.g. 1 โ†’ "1".

Quick start

DuckDB (Python, R or CLI). No download is needed:

import duckdb
con = duckdb.connect()
base = "hf://datasets/jamesjiadazhan/YRBSS/data"
con.sql(f"""
    SELECT Q2 AS sex, QN26, WEIGHT, STRATUM, PSU
    FROM read_parquet('{base}/2023/National_YRBS/XXH2023_YRBS_Data.parquet')
""").df()

# all state surveys of 2019 (2025 release)
con.sql(f"""
    SELECT sitecode, count(*) FROM read_parquet('{base}/2019/SADC_HS_State/sadc_2025_state_*.parquet')
    GROUP BY 1 ORDER BY 1
""").show()

Dictionary lookup:

con.sql(f"""
    SELECT variable, label, question_text, universe
    FROM read_parquet('{base}/2023/Dictionary/VARIABLES.parquet')
    WHERE file = 'XXH2023_YRBS_Data' AND variable IN ('Q26', 'QN26')
""").show()
con.sql(f"""
    SELECT code, description, published_unweighted_n, published_weighted_pct, n_in_file
    FROM read_parquet('{base}/2023/Dictionary/VALUES.parquet')
    WHERE file = 'XXH2023_YRBS_Data' AND variable = 'Q26'
""").show()

R (arrow + survey):

library(arrow); library(survey)
base <- "https://huggingface.co/datasets/jamesjiadazhan/YRBSS/resolve/main/data"
d <- read_parquet(file.path(base, "2023/National_YRBS/XXH2023_YRBS_Data.parquet"))
des <- svydesign(ids = ~PSU, strata = ~STRATUM, weights = ~WEIGHT, nest = TRUE, data = d)
svyciprop(~I(QN26 == 1), subset(des, !is.na(QN26)), method = "logit")   # 39.7% (37.7-41.7)

Hugging Face datasets (see Hugging Face subsets and splits for every subset):

from datasets import load_dataset
st = load_dataset("jamesjiadazhan/YRBSS", "sadc_2025_state", split="year_2019")

Survey design: required for every estimate

The national YRBS uses a three-stage cluster sample. Use the design for every estimate:

Files Weight Stratum PSU
National YRBS 1991โ€“2025 WEIGHT (Weight in 2003โ€“2007) STRATUM (Stratum in 2003โ€“2007) PSU
College 1995 WEIGHT STRATUM PSU
SADC (all) weight stratum PSU
  • In R: survey::svydesign(ids = ~PSU, strata = ~STRATUM, weights = ~WEIGHT, nest = TRUE, data = d). CDC uses SUDAAN with Taylor series linearization. Proportions with logit-transformed confidence intervals (svyciprop(method = "logit")) reproduce CDC's published intervals.
  • National weights are scaled so that they sum to the sample size of each year, with grade proportions matching national population projections.
  • State and district (SADC) surveys are separate samples, each with its own weights, strata and PSUs. Analyse each site separately or follow CDC's Combining YRBS Data Across Years and Sites. CDC advises against combining state with district data: they may overlap, and district data contain only urban students. Some states and districts are unweighted in some years and are then not in the SADC files at all.
  • Combining years: each year is an independent sample. Stack years and keep year-specific strata and PSUs. Nesting PSU within stratum within year is the safe choice, since stratum codes repeat across years. Keep each year's own weights. For trends, CDC uses logistic regression with linear and quadratic year terms on the SADC national file (Conducting Trend Analyses of YRBS Data).
  • CDC uses a minimum denominator of 30 for reporting.

metadata/YEARS.parquet lists, for each survey year, subset and release: the number of rows, files and sites; the design variable names as they appear in the files; and this guidance.

Validation

pipeline/05_validate.R writes these checks; all pass except the noted ones.

  • Row counts. Every national file has the number of usable questionnaires stated in its data user's guide (e.g. 20,103 in 2023 and 10,710 in 2025). The college file matches the sum of its published codebook frequencies.
  • Source lines. In every file, the year parts together hold every line of the CDC .dat file.
  • SADC national. For each year, the SADC national rows equal the national file rows, in both releases.
  • Keys. source_row is unique in every file. SADC record is unique within each file.
  • Unweighted frequencies. For every national year 1991โ€“2023, at least 99% of the published unweighted frequencies equal the counts in the data (100% in most years). Exception: in the 1995 college codebook 15 of 643 rows are misread from the PDF (Q13, Q14, Q33, Q50).
  • Weighted percentages. For 1999โ€“2023, at least 99% of the published weighted percentages are reproduced within ยฑ0.05 points. Exception: in 2001 the codebook's geography variables (GREG, METROST) do not match, although their unweighted counts do; the remaining 505 percentages match.
  • Published estimate. Persistent feelings of sadness or hopelessness, 2023 (QN26). The data give 39.7% (95% CI 37.7โ€“41.7) overall, 52.6% (50.1โ€“55.0) for female students and 27.7% (25.9โ€“29.6) for male students. These are identical to MMWR Suppl 2024;73(4) Table 2, using Taylor series linearization with logit confidence intervals.

Sources and processing

Part Source Processing
National YRBS, college CDC YRBSS data: ASCII .dat, SAS input/format programs, data user's guides Fixed-width read with the CDC SAS layout; values unchanged
Combined datasets (SADC) CDC YRBSS data: sadc_2025 (HS), sadc_2023 (HS, MS) Same; split by year
Dictionaries SAS input and format programs; codebook and appendices of the user's guides (pdftotext -layout) Parsed; see table above

CDC also distributes each file as a Microsoft Access database (.zip). Those copies hold the same data in another format and are not used. The SPSS syntax files are mirrored by the pipeline but are not needed.

Not included: the 2019 and 2021 SADC releases, which are superseded by later releases but may contain sites later withdrawn. Also excluded are YRBS data that CDC releases only on request through the data request form, and special studies without public files on the data page.

Licence and citation

  • YRBSS data and documentation: CDC materials on cdc.gov are in the public domain. CDC's Use of Agency Materials states: "Most of the information on the CDC and ATSDR websites is not subject to copyright, is in the public domain, and may be freely used or reproduced without obtaining copyright permission." The same page sets conditions: attribution to CDC, a non-endorsement disclaimer, no change to substantive content, and a statement that the material is available from the agency website at no charge. This card provides all four, and the data values are unchanged.
  • Cite: Centers for Disease Control and Prevention. Youth Risk Behavior Surveillance System (YRBSS) data (year of survey). Atlanta, GA: U.S. Department of Health and Human Services, CDC. https://www.cdc.gov/yrbs/data/index.html
  • The pipeline, dictionaries and layout of this compilation are released under CC0-1.0.
  • The data are de-identified, school-based survey responses. Do not attempt to identify any student or school.
Downloads last month
141