Objective:To compare the effectiveness of learning procedural skills on patients versus mannequins and models.Methods:Seventy four interns from two consecutive batches at the Department of Obstetrics and Gynaecolgy Unit-I at Jinnah Postgraduate Medical Center Karachi participated in the study between April and September 2014. Five basic skills; taking a cervical (Pap) smear, intrauterine contraceptive device insertion, manual vacuum aspiration, making/ suturing an episiotomy and active management of the third stage of labour were identified. Interns were randomly allocated to two training groups (Group-1 and 2 of thirty eight and thirty six trainees respectively), with Group-I received training on the five procedural skills on models and mannequins for four weeks while Group-II trained on patients initially. After an evaluation at four weeks the groups crossed over with a final evaluation at eight weeks. The evaluation was through identical objective structured assessment of technical skills on models and mannequins for both groups with standard checklists.Results:There was no significant difference in skills between the two groups at the four weeks assessment. However at the end of training, Group-1 trainees performed significantly better than Group 2 with higher overall tests scores (86.7 ± 2.7 versus 80.4 ± 4.8, p< 0.001). This difference was more marked in skills of intrauterine contraceptive device insertion, making and suturing an episiotomy and active management of third stage of labour.Conclusion:Our findings suggest that simulations using models and mannequins for developing procedural skills can be readily incorporated in training programs with potential benefits for teaching infrequently performed or more difficult procedures. Our data suggest potential benefits of initiation of trainings on simulations and mannequins followed by human subject exposure.