Three experiments were performed to investigate time-based retention performance in aversively motivated behavior. In Exp. 1 separate groups of rats received one-way active avoidance or yoked noncontingent footshock training at 4 pm either 0.25, 6, 12, 18, 24, or 30 hr prior to one-way active avoidance testing. The data revealed neither of the two most commonly observed retention functions (i.e., U-shaped “Kamin effect” or a multiphasic pattern). In Exp 2 the same procedure was employed except all subjects were trained at 9 am. A multiphasic pattern of retention performance for both active avoidance and noncontingent shock groups was observed. Experiment 3 employed a “truly noncontingent” control procedure. Results indicated that the retention functions observed in Exp 2 were primarily due to associative rather than nonassociative processes.