Preliminary data suggest there is a small and identifiable number of youths who are both violent offenders and mentally ill, a few States have established special programs for this population, and at least some of these programs have proven effective. In addition, issues which policy and program planners should address in developing these special programs have been identified. These issues include decision criteria, agency auspices, selections, service provider, and aftercare. For administrators and program planners to make informed decisions on the issues, and for this hard-to-treat population to receive the most appropriate services, more research is needed. A larger data base on both the youths themselves and the system's response must be developed before major progress can be made. Until such research is conducted, it is naive to expect juvenile corrections and mental health administrators to make any major improvements in the system's response to treating what is perhaps the hardest-to-treat population under State auspices. Footnotes and 20 references are included. (Author summary modified)