Objective To explore the relationship between different categories of internet addiction and depressive symptoms among university students using latent profile analysis and analyze the mediating effect of cognitive emotion regulation strategies between the two. Methods Random sampling was used to select 889 university students in five grades from four colleges and universities in Xinxiang City in January 2023 for the study. The General Demographic Information Questionnaire， internet Addiction Questionnaire， Cognitive Emotion Regulation Questionnaire （CERQ）， and Patient Health Questionnaire-9 （PHQ-9） were used to investigate internet addiction， depressive disorder， and cognitive emotion regulation strategies among university students. Mplus 8.3 was used to analyze the latent profile of internet addiction. Pearson correlation was used to analyze the correlation between internet addiction， cognitive emotion regulation strategies， and depressive disorder among university students. PROCESS procedure of SPSS 27.0 was used to test for the mediating role played in the relationship between internet addiction categories and depressive disorder and cognitive emotion regulation strategies. Results A total of 889 questionnaires were distributed and 802 were recovered， with an effective recovery rate of 90.21%. Among 802 university students， the total score of internet Addiction Questionnaire， adaptive cognitive emotion regulation strategy dimension score， non-adaptive cognitive emotion regulation strategy dimension score in CERQ， and PHQ-9 score were （44.33±13.03）， （66.20±11.48）， （42.71±9.51）， and （5.98±4.51）， respectively. Pearson correlation analysis showed that， internet addiction， adaptive cognitive emotion regulation strategies， non-adaptive cognitive emotion regulation strategies and depressive disorder see were all correlated， with statistically significant differences （r=-0.091-0.486； all P ＜ 0.01）. Latent profile analysis was conducted using each entry of the internet Addiction Questionnaire as an analytical indicator， and the number of categories was gradually increased starting from one latent category. A total of five category models were fitted， with model 3 being the best model. Naming each category based on the characteristics of the entries' responses， the results showed that there were three latent categories of university students' internet addiction， which were normal internet use group （36.3%）， low-level internet addiction group （43.3%）， and highlevel internet addiction group （20.3%）. Using normal internet use group as the reference group， both low-level internet addiction group and high-level internet addiction group positively predicted depressive disorder （β=0.33， P ＜ 0.001； β=0.75， P ＜ 0.001）. There was a mediating effect for both adaptive cognitive emotion regulation strategies and nonadaptive cognitive emotion regulation strategies between high-level internet addiction group and depressive disorder， with mediating effect values of 0.06 ［95%CI （0.016，0.120）］ and 0.39 ［95%CI （0.291， 0.491）］. There was a mediating effect of non-adaptive cognitive emotion regulation strategies between low-level internet addiction group and depressive disorder， with a mediating effect value of 0.22 ［95%CI （0.156，0.297）］. Conclusions There is a certain amount of internet addiction among university students. Between high-level internet addiction and depressive disorder， there are mediating effects of adaptive and nonadaptive cognitive emotion regulation strategies. Between low-level internet addiction and depressive disorder， there is a mediating effect of nonadaptive cognitive emotion regulation strategies.