Traditional anatomy education, which primarily relies on two-dimensional imagery, often struggles to effectively convey the complex spatial relationships of human anatomy. Virtual reality and three-dimensional (3D) anatomy models present a promising solution to these limitations. This study investigates the impact of integrating 3D anatomy models into a blended learning framework across pre-class, in-class, and post-class phases. A total of 169 medical students from Xinjiang Medical University were divided into three groups: a control group (Class A, = 57) following a traditional blended learning approach, and two experimental groups: Class B ( = 56), which incorporated continuous 3D model integration, and Class C ( = 56), which adopted a phased 3D model integration strategy. Learning outcomes and student satisfaction were assessed through formative evaluations, surveys, and statistical analyses. Our analytical framework employed dual statistical validation protocols: parametric testing via independent samples t-tests for normally distributed data and non-parametric verification through Mann-Whitney U tests for skewed distributions. Class B achieved higher scores than Class A across two assessment stages ( < 0.05). In pre-class evaluations, Class B ( = 56) scored 69.7 ± 7.5 compared to Class A's 63.8 ± 6.9 ( = 57). This performance gap persisted during in-class assessments, with Class B attaining 77.1 (± 8.7) against Class A's 70.8 (± 7.6). Prior to the intervention, Class C ( = 56) exhibited a mean score of 61.8 ± 6.1, which increased to 67.0 ± 6.7 post-intervention. The score gaps demonstrate the teaching method's effectiveness Class C demonstrated a statistically significant enhancement in pre-class assessment performance ( < 0.05) following the implementation of 3D anatomical modeling. However, no significant differences were observed among the groups in midterm or final exam scores ( > 0.05). Satisfaction scores in Class B were significantly higher than in Class A ( < 0.05), particularly in aspects of learning interest and teaching diversity. Class C also reported increased satisfaction in some dimensions after 3D model integration ( < 0.05). All survey instruments demonstrated good reliability (Cronbach's alpha > 0.7). In conclusion, while 3D anatomy models enhance student engagement, learning efficiency, and overall satisfaction, their effect on long-term retention and final exam performance remains limited. These findings underscore the need for a strategic approach to integrating 3D technologies in anatomy education to maximize their educational benefits.