An important development in social policy has been the appearance of prison reentry programs designed to help ex-offenders transition back into society when they leave prison, chiefly by promoting employment. Early evaluations of these programs show that they can promote employment and reduce recidivism, a rare success in the criminal justice field. It is reasonably clear from evaluations and field experience that these programs should require participation, provide case management, stress placement in available jobs over training, follow up with clients after job placement, and help them with other problems such as health and housing. However, many design questions remain unanswered. It is unclear whether the programs must create transitional jobs for their clients, versus placing them immediately in private employment. How long should transitional jobs last? What is the best form of sanctions to maintain participation in the program? What share of ex-offenders should best be assigned to work programs? How should programs best address drug dependency? Should they be residential? What is the role of contractors? What is the best interface with the child support system, which also seeks work programs to move low-skilled men into employment, so they can pay child support? In reentry panels that I organized at prior APPAM conferences, the emphasis was on the evaluation results showing impacts. The proposed roundtable will not repeat those findings but focus on the unresolved design issues. It will say more about how programs can best optimize their results. I will submit a list of the design issues to the participants before the conference and ask them to prepare responses. At the roundtable, I will summarize these issues, and then the speakers will address the most important of them, followed by discussion. Three of the participants are leading entrepreneurs in this field whose diverse views will help us resolve, or at least sharpen, these issues. They will speak for one program that is nonresidential and uses only private job placement (Lee Bowes, American Works), one that is both residential and uses transitional jobs (George McDonald, the Doe Fund), and one that is nonresidential and uses transitional jobs (Mindy Tarlow, Center for Employment Opportunities). Ron Mincy, our last speaker, has extensive field experience studying and advising fatherhood programs. He has well-formed views on the best shape of men’s work programs that deserve a larger audience. I proposed virtually this same panel for the 2011 conference. It was declined, leaving the conference with no panels on reentry programs. Meanwhile, the subject has assumed new urgency with the recent naming of a National Academy of Sciences committee on how to minimize incarceration. Reentry programs are critical to that goal, but there is little research bearing on these design and implementation issues. I am on this committee and thus will have an updated fix on the key issues by conference time. [471 words]