We tested the effectiveness of a prompt-and-praise procedure for increasing walking distance and independence in eight nursing home residents. Walking distance and method of ambulation were observed just prior to mealtimes; in addition, mobility and social interaction were time-sampled in the living areas throughout the day. Treatment was lagged in a multiple baseline design across lunch and dinner meals within subjects, and across subjects within each of three units. In the mealtime setting, two subjects began walking the maximum scored distance during baseline; the other six subjects showed a marked increase in walking beginning with the first meal in which the intervention was applied. Six of the eight subjects also progressed to more independent means of ambulation. Generalization of walking to the second meal was observed in all four of the subjects in whom this could be assessed. Generalization across subjects was not observed nor was generalization to the living areas. Staff successfully implemented the procedures in the mealtime settings and the effects were maintained at the 4-month follow-up.