BACKGROUND: HIV self-testing (HIVST) kits are common in key sexually active populations. Direct secondary distribution of HIVST kits (DSDHK) is effective in improving the uptake of HIVST. However, there are concerns about the various limitations of DSDHK, including limited geographic reach, payment problems, and need for face-to-face interactions. OBJECTIVE: In this study, we aim to evaluate the feasibility and characteristics of indirect secondary distribution of HIVST kits (ISDHK) via WeChat (distributing HIVST application links and follow-up HIVST kits to partners) among men who have sex with men (MSM). METHODS: From October 2017 to September 2019, an HIVST recruitment advertisement was disseminated on the WeChat social media platform to invite MSM to apply for the HIVST kits (referred to as index participants [IPs]). All MSM participants were encouraged to distribute the HIVST application link to their friends and sexual partners (referred to as alters) through their social networks. All the alters were further encouraged to continue distributing the HIVST application link. All participants paid a deposit (US $7), which was refundable upon completion of the questionnaire, and uploaded the test results via a web-based survey system. RESULTS: A total of 2263 MSM met the criteria and successfully applied for HIVST. Of these, 1816 participants returned their HIVST results, including 1422 (88.3%) IPs and 394 (21.7%) alters. More alters had condomless anal intercourse, a higher proportion of them had never previously tested for HIV, and they showed a greater willingness to distribute HIVST kits to their sexual partners (P=.002) than the IPs. After controlling for age, education, and income, the alters had a greater proportion of MSM who had never tested for HIV before (adjusted odds ratio [aOR] 1.29, 95% CI 1.00-1.68), were more willing to distribute the HIVST application link (aOR 1.71, 95% CI 1.21-2.40), had a lower number of sexual partners (aOR 0.71, 95% CI 0.57-0.90), and were less likely to search for sexual partners on the web (aOR 0.78, 95% CI 0.60-1.02) than IPs. In comparison, the rates of reactive HIVST results, conducting HIV confirmatory tests, HIV seropositivity, and initiation of HIV antiretroviral therapy were similar for IPs and alters. CONCLUSIONS: The ISDHK model of distributing HIVST application links among the MSM population via social media is feasible. The ISDHK model should be used to supplement the DSDHK model to enable a greater proportion of the MSM population to know their HIV infection status.