Token reinforcement procedures are used to increase appropriate behavior (e.g., completion of instructional tasks) and decrease problem behavior (e.g., aggression) in a wide array of clinical contexts. Token exchange-production schedules, which specify the number of tokens that must be accrued to produce the exchange schedule, are important components of token reinforcement programs. For example, accumulated schedules require learners to "bank up" multiple tokens before they can be exchanged, whereas distributed schedules allow learners to exchange tokens for backup reinforcers after just a single token (or a comparatively smaller range of tokens) is earned. Emerging research suggests that individual preferences for these different arrangements varies and is affected by many variables, including increasing work requirements (e.g., reinforcement schedule thinning). In the current study, we demonstrated that preferences for these schedules was stable across increasing work requirements and demonstrated that non-preferred (accumulated) schedule arrangements may support appropriate behavior and decrease problem behavior, even with very lean token-production schedules. These results have implications for practitioners who use token-based interventions and conduct schedule thinning for escape-maintained behavior.