Objective:To understand the characteristics of sexualized drug use (SDU) among men who have sex with men (MSM) in Chengdu and classify their behavioral patterns by applying the two-step cluster (TSC) algorithm and by exploring the association between SDU patterns and high risk sexual behaviors, in order to provide a reference basis for the development of HIV prevention interventions.Methods:Supported by an MSM-friendly organization in Chengdu from December 2021 to February 2022, MSM were recruited by on-site survey and peer referral to collect information on sociodemographics, SDU characteristics, sexual behaviors, STD diagnosis, and HIV status. TSC was performed to classify the characteristics of SDU, and the differences between groups were compared. Chi-square test and multiple binary logistic regressions were used to identify the relationship between SDU clusters and HIV-related risk sexual behaviors.Results:A total of 727 MSM were surveyed, and 39.8% (289/727) of the respondents reported SDU experience with a same-sex partner in the last six months. TSC clustered SDU-MSM into three behavioral pattern groups. There were significant differences in monthly income, types of drugs used for SDU, mode, frequency, polydrug use, multiple sexual partners, non-adherence to condom use, and group sex among the three groups of SDU-MSM ( P<0.05). Multivariate binary logistic regression analysis showed that the third category of SDU-MSM was 2.22 (95% CI: 1.06-4.66) times more likely than the first category not to use condoms consistently; the third and second categories were 2.82 (95% CI: 1.18-6.77) times and 8.78 (95% CI: 3.42-22.42) times. Conclusions:The prevalence of SDU among MSM in Chengdu was more than 1/3, and different SDU clustering pattern was associated with high-risk sexual behaviors. MSM with higher SDU frequency, drug use, and polydrug abuse are more difficult to adhere to condom use and more likely to have group sex and high risk sexual behaviors, increasing the risk of STD and HIV infection, and need to strengthen SDU surveillance and intervention.