We evaluated a commonly used component of brief time‐out, in which release from time‐out is delayed contingent on the occurrence of disruption. Data were collected for one normal and two mentally retarded children on time‐out‐producing behaviors (aggression and disruption) as well as delay‐producing behaviors during time‐out (loud vocalizations, out‐of‐chair, aggression, and disruption). The results of a combination ABAC reversal and multiple‐baseline design indicated that, under the conditions used in this investigation, both delay and no delay variations were effective in reducing the frequency of the target behaviors. Implications for the use of time‐out to reduce aberrant beha