The rapid decline in the welfare caseload remains a subject of keen interest to both policymakers and researchers. In this paper, I use data from the Survey of Income and Program Participation spanning the period from 1986 to 1999 to analyze how the economy, welfare reform, the earned income tax credit (EITC), and other factors influence welfare entries and exits, which in turn affect the caseload. I find that the decline in the welfare caseload resulted from both increases in exits and decreases in entries. Entries were most significantly affected by the economy, the decline in the real value of welfare benefits, and the expansion of the EITC. Exits were most significantly affected by the economy and federal welfare reform. Federal reform had its greatest effects on longer‐term spells of the type generally experienced by more disadvantaged recipients. Some out‐of‐sample predictions help explain the otherwise puzzling observation that, despite substantial increases in the unemployment rate since 2000, caseloads have remained roughly constant. © 2004 by the Association for Public Policy Analysis and Management