Reducing therapeutic restraint is a desirable outcome for programs that serve individuals who exhibit challenging behaviors. This study investigated the effects of modifying the criterion for release from therapeutic restraint on frequency and duration. Release from restraint was changed from a behavior-contingent criterion (restraint terminated when a specified duration of calm behavior was demonstrated) to a fixed-time criterion (restraint terminated when a predetermined duration elapsed independent of behavior) with one child and two adolescents with acquired brain injury and difficult-to-manage behaviors. For all three students, the duration of time they were exposed to therapeutic restraint decreased with the fixed-time release criterion, whereas the frequency of restraint did not change. Practical implications of these results and future research opportunities are discussed. (Contains 4 figures.)