Differential negative reinforcement of alternative behavior (DNRA) without extinction may be clinically useful when practical or clinical restrictions preclude the use of extinction. DNRA without extinction, when the magnitude of escape for alternative behavior and problem behavior is equal, has not resulted in successful treatment outcomes. DNRA without extinction, when discrepant escape durations for compliance and problem behavior are used, has been found successful for one participant with escape-maintained problem behavior. This finding has not been replicated. Therefore, the potential utility of using discrepant reinforcer magnitudes in the context of DNRA without extinction remains unclear. The present study sought to replicate previous research by assessing the utility of using discrepant reinforcer magnitudes for the compliance and problem behavior of children with autism in the absence of extinction. Results indicated that a large discrepancy in reinforcer magnitude resulted in shifts in response allocation from problem behavior to compliance, whereas moderate and equal discrepancies did not. These findings suggest that magnitude manipulations enhance the effects of DNRA without extinction and that the procedure may be clinically useful when extinction is not practical. These data were presented at ABAI 2015 and BABAT 2015.