The analysis compared patients who had stayed on the program at least 2 years, and those who had dropped out; outcome measures were against the patients' own baseline level of functioning. Comparisons on criminal activity showed significant differences in favor of patients who stayed in the program. The findings also revealed that these two groups had different criminal justice histories prior to program admission. Earnings data showed that the drug treatment program did not affect the number of patients employed, but did help patients increase their post-program earnings. Patient drug use showed anticipated decreases for patients who retained on program; variables associated with success included increased employment, decreased criminal activity, and other lifestyle changes. Stepwise multiple regression analysis confirmed these initial results. 1 chart, 5 tables, and 6 notes