As suggested by social learning theory, verbalization by observers has mediational functions and promotes the notable speed of the observational learning and long-term retention of modeled contents. The rehearsal of modeling stimuli by observers also plays an important role in the retention subprocess of observational learning. Two experiments were conducted to examine the effects of the verbalization, rehearsal by observers, and the difficulty of task on observational learning. Experiment I tested the relative contribution of the verbalization by observers and the difficulty of task on observational learning. The Ss (observers) were 44 undergraduate students. The observational learning task was a successive discrimination task with four dimensions. A 2×2×4 factorial design was introduced. The first factor was the verbalization by observers, that is, the Ss observe the modeling stimuli with guided verbalization or not. The second factor was the difficulty of the observational task (difficult or easy task). And the last factor was the learning phases of observers; Ss perform 20-successive-trials as one learning phase four times and among these learning phases model also exhibits 20trials. Experiment II tests whether rehearsal by observers facilitates the observational learning or not. 33 Ss, also undergraduate students, were assigned to one of four treatment conditions involving (a) verbalization plus rehearsal (b), verbalization, and (c) no treatment, Ss in the first group were asked to rehearse the modeling stimuli before their own trials and the latter groups were identical with two groups with difficult task in experiment I. The major findings of these experiments were as follows. The Ss, who learned the easy task, performed significantly faster and better than the Ss, who learned the difficult task. However, the latter Ss performed better, when they were assisted with guided verbalization of modeling stimuli. In the later learning phase the verbalization was more effective than in the earliest learning phas