On a ward for 24 psychogeriatric patients, 18 were randomly assigned to three groups with different experimental conditions. The first group was prompted and reinforced for different activities; the second was only prompted; and the third served as a control group. Activities, both the trained ones and others, were continuously monitored throughout the five week study. The results showed a clear increase in the trained activities after training. The prompting condition was effective, and the addition of reinforcement did not add any to the effects. There was no generalization to untrained activities, and the effects did not last at the 1 hr or 21 hr assessments, but was clearly visible at 5 min after training. Time seemed to be more important than place, i.e. the effect ceased with elapsed time, not because of changes in the situation.