OBJECTIVE: To evaluate HIV self-testing performance and results interpretation among female sex workers (FSWs) in Kampala, Uganda, who performed unassisted HIV self-testing. METHODS: In October 2016, 104 participants used an oral HIV self-test while under observation by research assistants. Participants were not assisted on HIV self-test use prior to or during testing, and were only given the manufacturer's pictorial and written instructions to guide them. Research assistants recorded if participants completed and/or had difficulties with steps in the HIV self-testing process on a prespecified checklist. Randomly drawn, used HIV self-tests were interpreted by FSWs. We calculated the concordance between FSWs' interpretations of self-test results with those indicated in the manufacturer's instructions. RESULTS: Only 33% (34/104) of participants completed all of the key steps in the HIV self-testing process, and the majority (86%, 89/104) were observed having difficulties with at least one of these steps. Misinterpretation of HIV self-test results were common among FSWs: 23% (12/56) of FSWs interpreted HIV-negative self-test results as HIV positive and 8% (3/37) of FSWs interpreted HIV-positive self-test results as HIV negative. The concordance between FSWs' interpretations of self-test results and that indicated in the instructions was 73% (95% CI 56% to 86%) for HIV-positive self-tests and 68% (95% CI 54% to 80%) for HIV-negative self-tests. CONCLUSIONS: FSWs in Kampala, who performed unassisted HIV self-testing, skipped steps in the HIV self-testing process and had difficulties correctly interpreting self-test results. Training on use and interpretation of HIV self-tests may be necessary to prevent errors in the HIV self-testing process and to avoid the negative consequences of false-positive and false-negative HIV self-test results among FSWs. TRIAL REGISTRATION NUMBER: NCT02846402.