Behavioral momentum theory posits a paradoxical implication for behavioral interventions in clinical situations using Differential Reinforcement of Alternative Behavior (DRA): When alternative reinforcers are presented within the same context as the problem behavior, the added reinforcers may decrease the frequency of the behavior but also increase its persistence when the intervention ends. Providing alternative reinforcers in a setting that is distinctively different from that in which the target behavior occurs may avoid or reduce this increase in persistence. The present experiment compared behavioral persistence following standard DRA versus DRA in a different context that was available after refraining from target behavior (Differential Reinforcement of Other Behavior, DRO). We arranged a human laboratory model of treatment intervention using computer games and token reinforcement. Participants were five individuals with intellectual disabilities. Experimental phases included (a) an initial multiple-schedule baseline with token reinforcement for target behaviors A and B, (b) an intervention phase with alternative reinforcement using a conventional DRA procedure for A and a DRO-DRA procedure for B, and (c) an extinction phase with no interventions and no tokens. Response rates as proportion of baseline in the initial extinction phase were greater for A than for B for three of five participants. Four participants whose response rates remained relatively high during the extinction phase then received a second extinction-plus-distraction test with leisure items available. Response rates were greater for A than for B in three of four participants. The results indicate that DRO-DRA contingencies may contribute to reduced post-intervention persistence of problem behavior.