Token economies incorporating differential reinforcement of other behavior and response cost have been shown to be effective for decreasing behavior. However, these token economies in the literature have differed in their findings of the effectiveness of and preference for differential reinforcement of other behavior and response cost. They have also differed in the way they implemented token economies including factors such as the immediacy of a consequence and the use of verbal feedback. Therefore, we examined two versions of response cost, one in which the contingency is experienced immediately and one in which it is experienced at end of an interval, with differential reinforcement of other behavior. We also examined these contingencies with the addition of verbal feedback and the participant preference for conditions. Results demonstrated that there was a faster decrease in errors during the response cost immediate conditions when compared to response cost delay and DRO conditions. All conditions resulted in fewer errors than the conditions without verbal feedback and preference was found to be variable. Response cost immediate resulted in the most raffle tickets earned during the no verbal feedback phases and most of the verbal feedback phases.