The history of unpaid work experience shows that work requirements for welfare recipients have often been proposed and attempted. The consistent trend has been to place greater emphasis on employability services. Studies of nine welfare-to-work programs providing unpaid work experience indicate that the level of participation is limited by factors related to program design, welfare dynamics, worksite capacity, and client characteristics and circumstances. Worksite surveys of attitudes toward unpaid work experience show most participants view it as better than welfare, but not as good as a real job. Three studies that isolated effects of the Community Work Experience Program suggest that unpaid work experience does not have a clear or consistent effect on employment, earnings, or welfare receipt. Other welfare-to-work programs have proven more effective in increasing employment and earnings and reducing use of welfare. Studies on cost effectiveness indicate that work experience appears to have relatively little net effect on the economic well-being of welfare recipients. For policymakers considering an expansion of unpaid work experience, major uncertainties remain about scale, both in terms of the administrative feasibility and the applicability of the experience of previous smaller programs. (Appendixes include more information on calculating costs and assessing cost effectiveness. Contains 60 references.) (YLB)