This research was performed to compare the effect of various levels of contextual interference (blocked, random and increasing practice schedules) on acquisition, retention and transfer of movement time error and cursor error in a complex motor – perceptual task. For this purpose, 30 female students (age range of 11 – 13 years) participated in a pretest (3 parts of 12 trials) and based on their pretest scores, they were divided into 3 homogeneous groups: blocked, random and increasing. The participants performed the guidance and click by mouse tasks in accordance with their groups. 4 days after the pretest, the participants performed 288 trials (12 trials in 18 parts) in the acquisition phase and after 24 hours, they participated in retention and transfer phases (12 trials in 3 parts). Movement time error (MTE) and cursor error (CU) were analyzed separately for acquisition, retention and transfer phases by analysis of variance 3×4 and Tukey post hoc test. The result showed a significant difference (P=0.00) in the movement time error between the acquisition phases with confidence level of 99%, but there was no significant difference (P=0.29) in the cursor error. Also, there was no significant difference among the three groups in the acquisition, retention and transfer.