September 2026
- First released value
- +4.2
- Original release date
- Oct 2, 2026
CPS unemployment rate; no revised value recorded.
DATA FILE · DF-004
What share of the civilian labor force is unemployed, and what does the denominator leave out?
+4.2 percent of the civilian labor force, age 16 and over
The September unemployment rate was 4.2 percent, compared with 4.1 percent in August. BLS described the rate as having changed little. The 0.1 percentage-point difference is distinct from a 0.1 percent relative change. No revised August rate is recorded in this file.
Open the primary data release (opens in a new tab)Values use the same unit shown above. A revision changes the estimate for an earlier period; it is not a new reporting period’s reading.
Read the original previous-period release (opens in a new tab)The share of the civilian labor force classified as unemployed under BLS definitions. The labor force comprises employed and unemployed people; people outside it are not in the denominator.
It does not by itself show whether a falling rate reflects hiring or people leaving the labor force. It is not the share of all adults without jobs, the payroll-job growth rate, or a complete measure of underemployment, wages or job quality.
Series LNS14000000: Current Population Survey, civilian unemployment rate, age 16 and over, seasonally adjusted. Unit is percent; there is no additional transformation. The rate is unemployed people divided by the civilian labor force, multiplied by 100. CPS is a household survey of people, distinct from CES payroll jobs. Annual seasonal-adjustment revisions and population-control changes can affect comparisons; retain original release vintages and identify any later revisions explicitly.
CPS unemployment rate; no revised value recorded.
August's original CPS rate was 4.1 percent and remains 4.1 in the September release; no revision is asserted.
Nov 6, 2026. The agency’s release calendar controls the schedule.
A Lower Unemployment Rate Can Hide a Weaker Job Market
Read the articleFound a measurement error or a newer primary release?
Send evidence or a correction →