Welfare caseloads have fallen dramatically in the last several years, raising questions about the economic well-being of those who left. In this paper we use administrative data from Wisconsin to identify those who left welfare and to provide information on their economic well-being. We provide a context for understanding post-welfare well-being by comparing welfare leavers under early Wisconsin reforms (those who left in the fourth quarter of 1995) with those who left under the later, more stringent TANF program (those who left in the fourth quarter of 1997). We also provide three-year outcomes for those who left under the early reforms to examine changes in economic well-being over time.We find substantially higher rates of exit in the later cohort. In both cohorts, about 70 percent of leavers have earnings in each quarter of the next year. Earnings are lower in the second cohort, a finding consistent with the hypothesis that the new welfare regime pushes people with fewer employment skills into the labor market where they accept lower-paying jobs. Multivariate analyses show that those with higher levels of human capital (education, employment experience, previous earnings) do better when they leave welfare. We measure post-exit income by adding earnings, cash assistance, Food Stamps, and the estimated EITC and subtracting estimated payroll and income taxes. This limited measure of net income neglects other income sources and employment-related expenses. We find that leavers have substantially higher earnings and EITC than they did prior to exit, but the decline in benefits outweighs these increases so that total measured net income in the year following exit is lower. In both cohorts less than one-third of the leavers have higher incomes than they did previously. Official poverty rates are high in both cohorts, over 60 percent, but are especially high in the later cohort. In the first cohort, earnings increase over the three-year period, benefits decline, and overall net income shows small increases.