Objective To analyze the influencing factors of self-directed learning ability of medical undergraduates, and to inform the improvement of the independent learning ability of medical students. Methods From July to August 2021, a total of 3 614 medical undergraduates from six medical schools in Jiangsu Province, Guangdong Province, Liaoning Province, and Heilongjiang Province were selected through stratified multi-stage sampling by questionnaire survey method, and the self-developed Medical Student Autonomous Learning Competency Scale was used to assess the influence of individual, school, and family factors on the self-directed learning ability of medical students. Two independent samples t-test or one-way ANOVA and differences in mean scores of independent learning ability between multiple groups were used; stepwise regression in multiple linear regression analysis was used to identify the factors affecting medical students′ self-directed learning ability. Results The overall mean score of medical students′ self-directed learning ability was (3.68±0.56). The results of the stepwise regression analysis are three-fold. At the individual level, male students had higher self-directed learning ability than female students (β=-0.04, P=0.016); first year students had higher self-directed learning ability than the other higher grades (β=-0.06, P<0.001); the higher the self-assessed health score (β=0.14, P< 0.001) and higher self-rated sleep quality scores(β=0.20, P<0.001) were associated with higher self-directed learning among medical students. At the family level, medical students with high family income had lower self-directed learning ability compared to those with poor family economic conditions (β=-0.04, P=0.022); non-rural-born medical students had higher self-directed learning ability compared to rural-born medical students(β=0.06, P<0.001); and those who received authoritative education had higher self-directed learning ability compared to neglectful education style (β=0.10, P<0.001). And at the school level, medical students with class officer or student government experience were more capable of independent learning ( β=-0.10, P<0.001). Conclusions The level of self-directed learning ability of undergraduate medical students is in the upper middle range. Medical students′ capacity for self-directed learning varies at multiple levels: individual, family, and school. Educational administrators need to pay special attention to the self-directed learning ability of female students, senior medical students, and medical students with poor sleep quality and physical health. Attention needs to be paid to the improvement of self-directed learning ability of medical students with economic difficulties and different educational and growth environments. Relevant departments are suggested to create more opportunities for students to participate in campus activities, and create a good learning environment and a harmonious campus atmosphere, so as to improve the self-directed learning ability of medical students as a whole. It is recommended that the relevant departments create more opportunities for students to participate in campus activities and create a favorable learning environment and a harmonious campus atmosphere, so as to improve the overall self-directed learning ability of medical students.