Time limits are a central component of recent welfare reforms and represent a substantial departure from previous policy. However, several recent studies suggest that they have had no effect on welfare use. In this paper I attempt to reconcile those findings with results from Grogger and Michalopoulos, who find time limits to have substantial effects that vary by the age of the youngest child in the family. Using data from the Current Population Survey, I obtain results similar to those of previous analysts when I estimate models that constrain the effects of time limits to be independent of age. When I allow for age dependence and employ controls for time-varying state-level unobservables that may be correlated with the timing of welfare reform, however, I find that time limits have negative effects on welfare use and that those effects are stronger, the younger the youngest child in the family. The estimates suggest that time limits may account for 16 to 18 percent of the recent decline in welfare use among female-headed families.