The U.S. added 29,000 nonfarm payroll jobs in September, while unemployment rose 0.1 percentage point to 4.2%, the Bureau of Labor Statistics reported. BLS described both measures as little changed. The previous two months were revised down by 60,000 combined, including July moving from a gain to a loss. That revision changes the summer comparison behind the September headline.

That leaves two questions behind the headline. How much did the latest report change the picture of hiring, and why can payrolls rise by 29,000 while the unemployment rate increases? The BLS release answers both, but it uses two separate surveys and each measures a different part of the labor market.

The revisions changed the summer comparison

The establishment survey estimate for July moved from a 21,000 increase to a 10,000 decline. August moved from a 162,000 increase to a 133,000 increase. BLS said revisions use additional reports from businesses and government agencies, plus recalculated seasonal factors. The revised numbers replace the earlier estimates; they do not count jobs that disappeared after the original releases.

September’s 29,000 increase follows those revised figures. BLS described both nonfarm payroll employment and the unemployment rate as having changed little during the month. Its 12-month average monthly payroll gain was 45,000, so September’s result was below that average. The release does not identify one cause for the slower monthly gain, and the two revised months do not by themselves establish what will happen next.

Payroll jobs and unemployment come from different surveys

The 29,000 figure comes from the establishment survey, which measures nonfarm jobs, hours and earnings by industry. BLS’s CES FAQ explains that each establishment reports its payroll jobs, so a person working for two different establishments can appear twice. The survey also excludes self-employment and some other work arrangements.

The 4.2% unemployment rate comes from the household survey. BLS’s CPS guidance classifies someone as unemployed when they have no job, are available to work and have actively looked for work in the prior four weeks, subject to the survey’s rules for people awaiting recall. BLS reported 7.1 million people unemployed in September, little changed from August. The rate has stayed between 4.1% and 4.3% since March.

The household survey also estimated that employment rose by 406,000 in September. That does not replace the payroll estimate. The surveys ask different questions, cover different units and can move differently from month to month. A one-month divergence is a reason to read both measures rather than combine them into one total.

BLS counted 1.9 million people as long-term unemployed in September, essentially unchanged from August. They represented 27.1% of everyone unemployed. That measure adds duration to the headline rate: it shows how many people had been out of work for at least 27 weeks, while leaving individual job searches and local hiring conditions outside the national summary.

A workplace manager greets arriving staff beside a separate home scene where an adult answers a telephone interview.
Concept illustration: payroll jobs count employer-reported jobs while unemployment is based on a household survey.

A few industries moved in opposite directions

Health care added 17,000 payroll jobs, below its average monthly gain of 33,000 over the prior year. Construction employment changed little at 11,000, near its 12-month average of 10,000. Financial activities lost 7,000 jobs and remained 129,000 below its May 2025 peak. Temporary-help services fell by 10,900.

Those figures help explain why the national total can look quiet even as some industries grow and others contract. BLS said employment changed little across all major industries overall. The release is a monthly snapshot.

What to watch next

Average hourly earnings for private nonfarm employees rose 5 cents to $37.81 in September, up 3.0% over the year. The average workweek stayed at 34.4 hours. These measures describe employees' pay and hours alongside the job count. They provide no forecast of household income.

The next Employment Situation release is scheduled for November 6. The September figure is an initial monthly estimate, while the July and August revisions are now part of the comparison. Taken together, the report shows limited payroll growth in September, unemployment within its recent range and a weaker summer than earlier estimates showed.