Skip to main content
v2026.11,986 entries · CC-BY 4.0

NHTS: National Household Travel Survey Data, Weights and Citation

NHTS explained: FHWA’s travel survey, 2022 sample and response rates, which weights to use, a margin-of-error example and an APA citation.

Written and maintained by CASRAI Editorial Board

Published

On this page

The NHTS is the National Household Travel Survey, the Federal Highway Administration’s (FHWA, part of the U.S. Department of Transportation) national inventory of daily personal travel in the United States. Households report the trips each member made over one 24-hour travel day, by purpose, mode, travel time, time of day and day of week, in urban and rural areas and across all modes. It is not the same thing as NHTSA, the National Highway Traffic Safety Administration, which is a separate agency.

What the NHTS is and who runs it

FHWA describes the NHTS (and its predecessor, the Nationwide Personal Transportation Survey, NPTS) as the nation’s inventory of daily travel, and says it has conducted the survey since 1969 and regards it as the authoritative source on the travel behavior of the American public. It is the only national source for trends in personal and household travel, covering daily non-commercial travel by all modes together with traveler, household and vehicle characteristics. FHWA lists Westat as survey contractor on its programme page; the 2022 data user guide was produced by Ipsos. The public data and tools are distributed from the Oak Ridge National Laboratory site, nhts.ornl.gov.

Survey years and what changed in 2022

Earlier waves are NPTS 1983, 1990 and 1995; the NHTS waves with public data are 2001, 2009, 2017 and 2022 (the 2022 wave is branded NextGen NHTS). The 2022 wave was fielded from January 2022 to January 2023, with completed households between 20 January 2022 and 19 January 2023, and travel days spread across all seven days of the week over the twelve months. Core data were released on 15 November 2023.

Two design changes matter for anyone comparing years. First, 2022 used one-stage data collection (households were not first recruited and then sent a diary, as in the earlier two-stage design). Second, 2022 drew an address-based sample from the U.S. Postal Service Delivery Sequence File, with most respondents reporting through a web tool, while 2017 offered online and phone options. The Bureau of Transportation Statistics summary of 2022 travel trends warns that this method change was accompanied by adjustments for comparability and that users should assess the effect on their own estimates. Treat a 2017-to-2022 change as a result that needs that caution, not as a plain time series.

A 2024 NextGen NHTS is under way: FHWA states that data collection began on 1 November 2024 with a sample of 7,500 households, running to November 2025, and that the programme is moving to a biennial survey with annual passive origin-destination data. Those 2024 design figures should not be mixed with the 2022 figures in this guide.

What is collected: trips, travel days and loop trips

The unit of observation is a trip taken by a household member aged 5 or older on an assigned travel day. The travel day runs from 4:00 AM to 4:00 AM the next day. Loop trips, which start and end at the same place, are captured and flagged (variable LOOPTRIP1). A household counts as complete when all members aged 5 and over in a household of one to three people have completed travel diaries, or at least 75 percent of members in a household of four or more. Trip purpose is carried in several variables, including WHYTO, WHYFROM, WHYTRP90, WHYTRP1S and TRIPPURP.

The four data files and how they join

The 2022 public data come as four files, available as CSV, SAS and SPSS. The keys tell you how to link them.

File One row is Key variables
HOUSEHOLD A completed household HOUSEID
VEHICLE A household vehicle HOUSEID + VEHID
PERSON A household member HOUSEID + PERSONID
TRIP One trip by a member on the travel day HOUSEID + PERSONID + TRIPID

Join trips to persons on HOUSEID and PERSONID, and either of them to households on HOUSEID. Because the person key is the pair HOUSEID + PERSONID, join on both variables rather than on PERSONID alone.

Sample size and response rates

The 2022 sample was stratified by Census division and urban or rural area. The national sample produced 7,893 completed households; add-on samples bought by states and regional agencies (through FHWA’s pooled-fund arrangement) brought the total to 27,290. The user guide states margins of error of no more than 5 percent at the 95 percent confidence level for the national sample. The American Association for Public Opinion Research (AAPOR) 9th edition rates reported in the guide are 20.4 percent for the profile response rate (RR4) and 10.8 percent for the overall response rate (RR3). The data table below sets the 2022 sample against earlier waves.

The national share of the 2022 total is 7,893 / 27,290 = 28.9 percent. Anyone producing estimates for a small area should expect to rely on add-on households or accept wide intervals; see the sampling concepts in multistage sampling, stratified sampling and cluster sampling.

Which weight to use

NHTS weights reflect selection probability, eligibility, nonresponse and undercoverage. Estimates from unweighted counts can differ greatly from weighted ones, so never tabulate the raw files without a weight. The choice depends on the unit and the days you want to describe:

Unit of analysis 7-day (whole year) 5-day (weekday) 2-day (weekend)
Household WTHHFIN WTHHFIN5D WTHHFIN2D
Person WTPERFIN WTPERFIN5D WTPERFIN2D
Trip WTTRDFIN WTTRDFIN5D WTTRDFIN2D

Vehicle records use the household weight. The user guide’s own recipe for a rate such as trips per household by region is a ratio of weighted sums: the sum of trip weights divided by the sum of household weights, with both sums taken within the region.

Worked example: estimate, margin of error and confidence interval

The NHTS Data Explorer, FHWA’s online tabulation tool for the 2022, 2017, 2009 and 2001 data, applies the weights automatically and returns three tables for each query: the weighted estimate, the unweighted sample size and the margin of error (MoE) at 95 percent confidence. The interval is the estimate plus or minus the MoE. The Data Explorer user guide gives two examples that show how much the interval can vary.

Example 1, a well-supported estimate. Total trips by workers: 136,981.07 million, MoE 6,951.16 million.

  • Lower bound: 136,981.07 – 6,951.16 = 130,029.91 million
  • Upper bound: 136,981.07 + 6,951.16 = 143,932.23 million
  • Relative MoE: 6,951.16 / 136,981.07 = 5.1 percent

Example 2, a thin cell. Medical or dental person trips for one-person households: 1,193,126,572.21, MoE 516,110,885.15.

  • Lower bound: 1,193,126,572.21 – 516,110,885.15 = 677,015,687.06
  • Upper bound: 1,193,126,572.21 + 516,110,885.15 = 1,709,237,457.36
  • Relative MoE: 516,110,885.15 / 1,193,126,572.21 = 43.3 percent
Relative MoE Reading
Around 5 percent (Example 1) The interval is narrow relative to the estimate; the estimate is usable for comparison between large groups.
Above 40 percent (Example 2) The interval spans a wide range; the point estimate alone is not a reliable basis for a statement.

The Data Explorer guide advises checking cells with fewer than 30 unweighted observations and collapsing categories when a cell is thin. For general background see the dictionary entry on confidence interval.

Assumptions and common mistakes

  • Using the wrong weight. A weekday question needs the 5-day weight, a weekend question the 2-day weight, an annual question the 7-day weight.
  • Treating the add-on samples as the national sample. The national sample is 7,893 households; the 27,290 total includes add-ons.
  • Reading 2017 versus 2022 as a clean trend. The mode of collection and sampling design changed.
  • Ignoring nonresponse. An overall response rate of 10.8 percent (RR3) is why the weights adjust for nonresponse and undercoverage; they do not remove every possible bias.
  • Splitting too finely. Cells with under 30 unweighted observations produce wide intervals, as Example 2 shows.

How to calculate trips per household in Excel, R or Python

These snippets compute the weighted ratio described above. File names are whatever you download; the weight variables are those named in the 2022 user guide.

Excel (sheets named Trip and Household, with the weight columns in column C of each sheet):

=SUM(Trip!C:C)/SUM(Household!C:C)

R

trip <- read.csv("trip.csv") hh   <- read.csv("household.csv") sum(trip$WTTRDFIN) / sum(hh$WTHHFIN)

Python (pandas)

import pandas as pd trip = pd.read_csv("trip.csv") hh = pd.read_csv("household.csv") ratio = trip["WTTRDFIN"].sum() / hh["WTHHFIN"].sum() person_trips = trip.merge(pd.read_csv("person.csv"), on=["HOUSEID", "PERSONID"])

To produce the figure by group, filter both files to the group before summing. Variance estimates and the exact MoE are provided by the Data Explorer; the simple ratio above gives a point estimate only.

How to cite the NHTS (APA example)

APA 7th edition data-set pattern, with the agency as author and the release year of the data you used:

Federal Highway Administration. (2023). 2022 NextGen National Household Travel Survey [Data set]. U.S. Department of Transportation. https://nhts.ornl.gov

State which weight and which file you used in the methods section so the estimate can be reproduced.

FAQ

What does NHTS stand for?

National Household Travel Survey, run by FHWA. It is unrelated to NHTSA, the National Highway Traffic Safety Administration.

How often is the NHTS conducted?

Past public waves are 2001, 2009, 2017 and 2022, after NPTS waves in 1983, 1990 and 1995. FHWA says the programme is moving to a biennial survey.

What is the latest NHTS?

The most recent data release described here is the 2022 NextGen NHTS Core Data (15 November 2023). The 2024 NextGen NHTS began collecting data on 1 November 2024.

Is NHTS data free?

The data and the Data Explorer are published publicly at nhts.ornl.gov. Agencies that want a larger sample for their own area can buy add-on samples through FHWA’s pooled-fund Study 651.

Can I compare 2017 and 2022 results?

Only with care: the collection design changed from two-stage to one-stage and the mode mix shifted toward the web, and users are advised to assess the effect on their estimates.

Sources

  • Federal Highway Administration, National Household Travel Survey programme page, https://www.fhwa.dot.gov/Policyinformation/nhts.cfm (read 2026-11-13).
  • Federal Highway Administration, NextGen NHTS data page, https://www.fhwa.dot.gov/policyinformation/nextgen-nhts-data.cfm (read 2026-11-13).
  • Ipsos, 2022 NextGen NHTS Data User Guide V2 (rev. February 2024), https://nhts.ornl.gov/assets/2022/doc/2022%20NextGen%20NHTS%20User’s%20Guide%20V2_PubUse.pdf (read 2026-11-13).
  • FHWA, 2022 NHTS Data Explorer User Guide, public version, https://nhts.ornl.gov/assets/2022/doc/DE_public.pdf (read 2026-11-13).
  • Bureau of Transportation Statistics, Summary of Travel Trends: 2022 NHTS, https://rosap.ntl.bts.gov/view/dot/73764 (abstract read 2026-11-13).

Data: NHTS completed-household samples by survey year, 2001 to 2022

Completed households in the national sample and in state or regional add-on samples for four National Household Travel Survey waves, compiled from Table 1 of the 2022 NextGen NHTS Data User Guide.

Period covered: 2001/2022. Coverage: United States.

NHTS completed-household samples by survey year, 2001 to 2022
Survey yearNational sample households (households)Add-on sample households (households)
2022789319397
201726099103597
200925510124637
200126038approximately 40,000

Sources

Compiled by CASRAI from the sources above. Free to reuse under CC BY 4.0 with attribution.

Follow CASRAI

Research-administration guidance, standards updates and independent tool reviews.

Ask CASRAI · free to try

Ask about NHTS: National Household Travel Survey Data, Weights and Citation

2 free questions a day, no account.

An AI assistant for research: funder rules, compliance, methods and publishing. It links the sources behind its answers and labels answers that come from a web search.

Answers draw on CASRAI's guides and dictionary plus the federal and funder documents we index: Federal Register, Grants.gov, Regulations.gov and UKRI.

Works on this site and inside Claude, Cursor and the AI tools you already use.

Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays open to read, with no account and no card.

Referenced across the research world

University of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoUniversity of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logo
  • University of Cambridge logo
  • Columbia University logo
  • Crossref logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo

View CASRAI adoption →