This investigation provides a preliminary examination of the difference between programmed and obtained reinforcement rates and its potential influence during treatment of aggression in a natural setting. Following a functional analysis that suggested that the aggression of a boy with autism was negatively reinforced, intervention was implemented by the boy's mother. Concurrent fixed-ratio (FR) 1 FR 1 schedules of escape were arranged for manding and aggression. When mands failed to compete effectively with aggression, obtained reinforcement ratios were calculated; these indicated that obtained reinforcement varied from the programmed schedule for aggression but not for mands. Increasing the rate of prompts for mands resulted in an increase in mands and a decrease in aggression to near-zero levels.