What is NFP — Nonfarm Payrolls?
Nonfarm payrolls (NFP) count the number of paid U.S. workers excluding farm employees, private household workers, and non-profit employees. The monthly change in NFP — reported the first Friday of each month alongside the unemployment rate — is the single most market-moving labor-market release.
Live data: All Employees, Total Nonfarm
Federal Reserve Economic Data (FRED) — · Monthly · 147 observations
Most recent observation: 158,637 Thousands of Persons as of March 1, 2026.
Understanding NFP
Nonfarm payrolls count the number of paid U.S. workers excluding farm employees, private-household workers, and nonprofit employees. The monthly change in NFP — released the first Friday of each month — is arguably the single most market-moving U.S. data release. Treasury yields, equities, and the dollar all move within seconds of the 8:30 a.m. ET print.
NFP comes from the Current Employment Statistics program — an establishment survey of about 145,000 businesses and government agencies covering roughly 700,000 worksites. Actual payroll records are reported, making it a more reliable count than the household survey for total employment level.
The headline is the seasonally adjusted monthly change. Analysts also watch the participation rate, average hourly earnings, the unemployment rate, and the diffusion index (share of industries adding jobs). Revisions are substantial: the first print is often revised significantly in subsequent months as late-filing responses come in.
How NFP is calculated
BLS surveys about 145,000 employers each month about their paid employment in the week containing the 12th. Responses are weighted to population estimates, seasonally adjusted, and benchmarked annually against unemployment-insurance records. The headline figure is the net change in total nonfarm employment.
Historical context
Average monthly NFP growth has been about 170,000 in healthy expansion periods. The April 2020 COVID collapse destroyed 20.5 million jobs in a single month — the largest drop on record. The subsequent rebound averaged 1 million+ per month in several stretches. Recession troughs typically show NFP losses of 200,000–800,000 per month at the peak.
Frequently asked questions
Why are the payroll and household surveys different?
Different sample frames and purposes. The payroll survey counts payroll records at businesses (larger sample, tighter methodology for employment level). The household survey counts individuals by employment status (better for unemployment rate and self-employment). They usually agree on direction but can diverge at turning points.
Why are NFP revisions so large?
The first print is based on incomplete survey responses — typically about 70% of the sample. As more responses come in, the estimate is revised in the next two monthly releases. Annual benchmarking to unemployment-insurance records can produce additional large revisions each February.