An analysis of tasks in a vocational training program was performed by a human factors specialist. Mentally disturbed patients enrolled in the program rated the tasks for importance, difficulty, and frequency of performance. Subsequent to the rating, tasks were assigned a numerical value or points. Individuals then selected tasks based on their own goals and standards. A daily log of the selected work samples was kept by patient self-report. This procedure was conceptualized as a self-reinforcement which was intrinsicallly related to the log keeping process. In addition, multiple generalized reinforcers were used such as need for approval and increased pay for productivity. When patients did well, i.e., made a high score, they were given higher wages and approval. Conversely, when they did poorly, i.e., made a low score, disapproval and criticism were expressed. In either case, daily feedback was provided along with a series of tasks graduated in difficulty, importance and frequency. This patterning of tasks was made possible by the “a priori” usuage of learning principles described by B.F. Skinner. Task analysis and work measurement techniques provided the means to set goals at the start of treatment. This systematic rendering of behavioral engineering in a work setting provided an interdisciplinary approach for changing coping processes in the mentally disturbed.