This study evaluated effect of mental rotation (MR) training on learning outcomes and explored effectiveness of teaching via three-dimensional (3D) software among medical students with diverse spatial intelligence. Data from n = 67 student volunteers were included. A preliminary test was conducted to obtain baseline level of MR competency and was utilized to assign participants to two experimental conditions, i.e., trained group (n = 25) and untrained group (n = 42). Data on the effectiveness of training were collected to measure participants' speed and accuracy in performing various MR activities. Six weeks later, a large class format (LCF) session was conducted for all students using 3D software. The usefulness of technology-assisted learning at the LCF was evaluated via a pre- and post-test. Students' feedback regarding MR training and use of 3D software was acquired through questionnaires. MR scores of the trainees improved from 25.9±4.6 points to 28.1±4.4 (P = 0.011) while time taken to complete the tasks reduced from 20.9±3.9 to 12.2±4.4 minutes. Males scored higher than females in all components (P = 0.016). Further, higher pre- and post-test scores were observed in trained (9.0±1.9 and 12.3±1.6) versus untrained group (7.8±1.8; 10.8±1.8). Although mixed-design analysis of variance suggested significant difference in their test scores (P < 0.001), both groups reported similar trend in improvement by means of 3D software (P = 0.54). Ninety-seven percent of students reported technology-assisted learning as an effective means of instruction and found use of 3D software superior to plastic models. Software based on 3D technologies could be adopted as an effective teaching pedagogy to support learning across students with diverse levels of mental rotation abilities.