Fifty-two "hard core" mental patients who had been receiving psychotropic drugs were assigned to one of four subgroups, equated on descriptive variables, level of functioning, and bizarre behavior. All patients then participated in one of two social-environmental treatment programs, with one subgroup in each program receiving drug continuation in different capsule form, and the other receiving abrupt withdrawal with placebo substitution. Relative and interactional effects of drug conditions were evaluated 17 weeks after drug changes with concomitant exposure to active treatment programs, without patients, treatment staff, or assessors being aware that a drug study was underway. Except in early stages of treatment, where drugs interfered with improvement, continued low-dosage maintenance chemotherapy as prescribed in practice failed to produce direct effects or to contribute to responsiveness to the environmental programs.