To date, no randomized controlled trials have specifically addressed behavior changes after HIV self-testing (HIVST) among transgender women. This study aims to evaluate the effects of HIVST on changes in HIV testing behavior, frequency of condomless sex, and partner numbers among transgender women in China. Participants were recruited from 2 Chinese cities using both online and offline methods. Transgender women were randomly assigned to receive an HIVST intervention. Data from the previous 3 months were collected at baseline, 3 months, and 6 months. The primary outcome was the mean change in the number of HIV tests among transgender women during the 6-month follow-up. An intention-to-treat analysis was conducted. The statistical analysis used analysis of covariance and linear mixed-effects models. From February to June 2021, and 255 transgender women were recruited, of which only 36.5% (93/255) had a steady job, and 27.1% (69/255) earned less than US $414.9 of income per month. They were randomly assigned to the intervention (n=127) and control (n=128) groups. At 6 months, the mean number of HIV tests was 2.14 (95% CI 1.80-2.48) in the intervention group and 1.19 (95% CI 0.99-1.40) in the control group (P<.001), with increases of 0.84 (95% CI 0.54-1.14) and 0.11 (95% CI -0.19-0.41) over 6 months, respectively. The net increase was 0.73 (95% CI 0.31-1.15; P<.001), with a similar adjusted result. No significant differences in the frequency of condomless sex or partner numbers were observed between the 2 groups. HIVST is an effective strategy for enhancing regular HIV testing behavior among transgender women in China. This strategy should be combined with measures to address the financial vulnerability of the transgender women community to reduce subsequent risk behaviors, including condomless sex. Chinese Clinical Trial Registry ChiCTR2000039766; https://www.chictr.org.cn/showproj.html?proj=61402.