The effectiveness of programmed instruction with operant training (P.I.O.T.) in the language rehabilitation of severe aphasics was evaluated by comparison with an attention-placebo control treatment (A.P.C.) in which non-verbal tasks were administered. Patients were divided into two groups and received first either 4 weeks of P.I.O.T. or of the A.P.C. condition. Both groups then received 4 weeks of the alternative treatment. Both groups also received some speech therapy throughout. Only two out of 24 assessments showed differences, both indicating that the A.P.C. with non-verbal practice improved abilities more than P.I.O.T. Over the total 8-week period, significant improvements in abilities occurred on more than one-half of the measures. It is concluded that rehabilitation therapy by means of programmed instruction with operant training failed to produce improvement in the course of recovery of language or related aspects of communication in severely aphasic patients.