It has recently been reported that selective REM sleep deprivation (REMD) in college students results in memory impairment of the application of a set of rules in a logic task, but not recall of a paired associate task. The present experiments were designed to examine the effects of Total Sleep Deprivation (TSD) and (REMD) following acquisition of a pure motor task, the pursuit rotor. In Experiment 1, subjects (N = 90) were exposed to TSD for one of several nights following training. Results showed that TSD on the same night as training resulted in poorer performance on retest one week later. In Experiment 2, subjects (N = 42) were exposed to various kinds of sleep deprivation on the night of task acquisition. One group was subjected to REMD. Other groups included a non-REM awakening control group (NREMA), a TSD group, a normally rested Control group and a group allowed the first 4 h of sleep in the night before being subjected to TSD (LH - TSD) for the rest of the night. Results showed the REMD and Control groups to have excellent memory for this task while the TSD and LH - TSD subjects had significantly poorer memory for the task. The NREMA group showed a slight, but not significant deficit. It was concluded that Stage 2 sleep, rather than REM sleep was the important stage of sleep for efficient memory processing of the pursuit rotor task.