Abstract: Objective To preliminarily evaluate “one-on-one mentorship system” teaching model in the "4+4" pilot class at Peking Union Medical College Hospital and compare it with the "group rotation system" teaching model to preliminarily evaluate its application effectiveness. Methods Based on the curriculum design and actual clinical situations, the "group rotation system" and “one-on-one mentorship system” were implemented in 2020 and 2021 cohorts of students respectively in the "4 + 4" pilot program at Peking Union Medical College Hospital. The teaching effects of these two different models were preliminarily evaluated. Results During the internship under the “one-on-one mentorship system” in the 2021 cohort, 18 students (62.1%) were very satisfied with their internship experience, 8 (27.5%) were satisfied, and 3 (10.3%) had an average experience. Regarding knowledge acquisition during the internship, 13 students (44.8%) believed they had gained abundant knowledge, 11 (37.9%) had an improvement in clinical insight, and 5 (17.2%) considered their learning gains as acceptable. The results of the students' internship forms indicated that 7 students (24.1%) encountered technical issues, 2 students (6.9%) faced challenges in applying professional knowledge, and 2 students (6.9%) had difficulties in doctor-patient communication. Among the tutors, 10 (17.2%) were very satisfied with the completion of the students' internship forms, 22 (37.9%) were satisfied, 22 (37.9%) expressed relatively low satisfaction, and 4 (7.0%) were dissatisfied. In terms of students' internship performance, 12 tutors (20.7%) were very satisfied, and 44 (75.9%) were satisfied. The theoretical examination scores of the 2021 cohort upon leaving the department were significantly lower than those of the 2020 cohort, with a statistically significant difference (72.6 ± 7.9) points vs. (80.1 ± 8.4) points, P=0.001. The excellent rate in the theoretical examination upon leaving the department was higher in the 2020 cohort than in the 2021 cohort, and the difference was statistically significant (P=0.004). The pass rates of the 2021 and 2020 cohorts were 96.6% and 100% respectively, no statistically significant difference was observed between the two cohorts (P=0.537). Conclusion Each of the two teaching models has its own strengths. The "group rotation system" requires students to participate in clinical work as resident physicians, enabling them to encounter a wider range of diseases, which easily stimulates thinking about practical issues, leading to a greater desire to master basic knowledge during post-class review and helps students broaden their knowledge base. The “one-on-one mentorship system”, as a task-oriented teaching model focuses on the completion of pre-set learning tasks, and allows students to have more direct contact with clinical practice, playing a crucial role in shaping their future learning plans and professional identity.