I document evidence of a systematic bias in the widely-used estimate of annual hours worked using the Current Population Survey. The bias stems from multiplying the number of weeks worked by the median – instead of the mean – of weekly working hours. Theoretical and empirical evidence suggests that this bias leads to an overestimation of annual hours. •I document evidence of a systematic bias in the widely-used estimate of annual hours worked using the Current Population Survey.•The bias stems from multiplying the number of weeks worked by the median – instead of the mean – weekly working hours.•Theoretical and empirical evidence suggest that this bias can lead to an over-estimation of annual hours.