Objective To investigate the relationship between circadian rhythm type and sleep quality among university students, and to clarify the mediating roles of trait anxiety and sleep beliefs and attitudes. Methods An online survey was conducted among university students at a medical school by convenience sampling method. The circadian rhythm types, trait anxiety levels, sleep beliefs and attitudes, and sleep quality were evaluated by the morning and evening questionnaire-5 (MEQ-5), trait anxiety inventory (T-AI), brief version of dysfunctional beliefs and attitudes about sleep scale-16 (DBAS-16), and Pittsburgh sleep quality index (PSQI), respectively. Pearson correlation analysis and mediation analysis were employed to examine the relationships between circadian rhythm types, trait anxiety levels, sleep beliefs and attitudes, and sleep quality. Results A total of 238 questionnaires were collected and 233 were valid, with an effective rate of 97.9%. Pearson correlation analysis revealed significant correlations between circadian rhythm type, trait anxiety level, sleep beliefs and attitudes, and sleep quality (all P＜0.01). The mediating effect was examined with circadian rhythm type as independent variable, sleep quality as dependent variable, and trait anxiety level and sleep beliefs and attitudes as mediating variables. The mediating effect of the trait anxiety level (MEQ-5→T-AI→PSQI) accounted for 30.48% of the total effect (r＝－0.128, 95% confidence interval［CI］ －0.211 to －0.066). The mediating effect of the sleep beliefs and attitudes (MEQ-5→DBAS-16→PSQI) accounted for 12.62% of the total effect (r＝－0.053, 95% CI －0.106 to －0.014). The chain mediating effects of the trait anxiety level and sleep beliefs and attitudes (MEQ-5→T-AI→DBAS-16→PSQI) accounted for 11.19% of the total effect (r＝－0.047, 95% CI －0.085 to －0.022). Conclusion The circadian rhythm type of university students can negatively predict their sleep quality, and the trait anxiety level and sleep beliefs and attitudes play mediating roles.