The research was designed to validate procedures to teach apartment upkeep skills to severely handicapped clients with various categorical disabilities. Methodological features of this research included performance comparisons between general and specific task analyses, effect of an impasse correction baseline procedure, social validation of training goals, natural environment assessments and contingencies, as well as long‐term follow‐up. Subjects were taught to perform upkeep responses on their air conditioner‐heating unit, electric range, refrigerator, and electrical appliances within the context of a multiple‐probe across subjects experimental design. The results showed acquisition, long‐term maintenance, and generalization of the upkeep skills to a nontraining apartment. General task analyses were recommended for assessment and specific task analyses for training. The impasse correction procedure generally did not produce ac