Objective: To examine the association between a dual-supervisor training model and self-reported research training needs. Methods: A quasi-experimental study with natural grouping was adopted. A total of 114 clinical professional degree postgraduates (Grade 2023) from a hospital affiliated to a university in Qingdao were enrolled. According to the training plans established at enrollment, 57 students were assigned to the dual-supervisor group and 59 to the single-supervisor group. Finally, 56 students in the dual-supervisor group and 58 in the single-supervisor group completed the study. Questionnaire surveys on research needs were conducted before (March 2024) and one year after (March 2025) the supervisor guidance. Chi-square test, Fisher’s exact test, and McNemar test were used to analyze the differences between and within groups. Results: The two groups were comparable at baseline (P > 0.05). After one year of guidance, the dual-supervisor group reported lower levels of unmet training needs than the single-supervisor group in research process cognition (25.0% vs 48.3%, χ² = 6.892, P = 0.009), statistical software application (41.1% vs 67.2%, χ² = 7.562, P = 0.006), scientific graphing ability (39.3% vs 65.5%, χ² = 8.247, P = 0.004), and academic presentation ability (32.1% vs 55.2%, χ² = 6.215, P = 0.013). The dual-supervisor group showed significant reductions in self-reported training needs in multiple indicators before and after the intervention (P < 0.01), while the single-supervisor group showed no significant improvement (P > 0.05). Conclusion: The dual-supervisor training model was associated with lower self-reported unmet research training needs among clinical professional degree postgraduates, compared with the single-supervisor model in several domains, and is worthy of further promotion and improvement.