A single subject, multiple baseline design was used to examine the efficacy of a treatment package consisting of relaxation, reinforcement, and cognitive self-instruction on the reduction of severe nighttime fears in six children. Children with persistent fears of the dark, which resulted in disruption of their families' nighttime routine, and their parents participated in a five-week program consisting of three training sessions and nightly home exercise. A multimodal assessment was conducted through use of self-report, parental report, and behavioral measures. Results show amelioration of the disruptive bedtime behavior of five of the six children. However, results of the multiple baseline analysis indicate changes in the children's behavior were not directly related to treatment. For children with extended baselines, improvement preceded the implementation of treatment. The implications for these findings on fear of the dark and general fears in children are discussed. / M.S.