The aim of this study was to design a low-cost three-dimensional (3D) laparoscopic simulator and validate its training effectiveness. We designed a low-cost 3D laparoscopic simulator using magnifying glass and cardboard box. Thirty-two laparoscopic novices were randomly divided into 3D group and two-dimensional (2D) group. The 3D group was trained on 3D simulator four times with 24 hours interval, and the 2D group was trained on 2D simulator. Five standardized laparoscopic tasks were performed by novices in each training. In the second part, subjects were transferred to the opposite simulator for one test after 24 hours of the fourth training. The completing time and errors for each task were recorded to assess the construct validity of simulator. Finally, the face validity and the content validity were evaluated through a closed-ended questionnaire. There was no significant difference between the two groups in demographic or psychometric variables (P > .05). Compared with the 2D group, novices using 3D simulator had a better performance in five laparoscopic tasks, including a faster completing time (P < .001) and lower errors during training (P < .05). Additionally, the increased laparoscopic skill involved with our 3D simulator could be transferred to subsequent performance in 2D simulator (P < .05). Meanwhile, the score of face validity and content validity in our 3D simulator was significantly higher than that in 2D simulator (P < .05). Our 3D laparoscopic simulator effectively improved laparoscopic skills of novice surgeons, suggesting that the low-cost 3D simulator had satisfactory performance to satisfy requirement for novice training.