BACKGROUND: In Uganda, HIV testing rates are approximately 90% among women in antenatal care, with male rates much lower. The World Health Organization has recommended HIV self-testing (HIVST), and one promising model is for women in antenatal care to deliver HIVST kits to their male partners. We investigated the impact of this model on male partner testing rates. SETTING: Three high-volume antenatal clinics in central Uganda. METHODS: We implemented a cluster-randomized controlled trial comparing standard of care to intervention, with the primary outcome of self-reported male partner HIV testing. Women and male partners were followed at 1 and 3 months. We used unadjusted analyses and log-linear models with an intent-to-treat approach accounting for clustering. RESULTS: Study coordinators randomized 1514 women (777 intervention and 737 control). Baseline characteristics were balanced across arms with mean age (SD) of 25.2 (5.5) years and >44% with secondary education or higher. More male partners tested for HIV in intervention [576/746 (77.2%)] versus control [264/709 (37.2%)], P < 0.01. We identified 34 HIV-positive men in intervention versus 10 in control, with 6/26 (23%) and 4/6 (67%), respectively, reporting linking to care. CONCLUSIONS: Our results demonstrate an enormous increase in self-reported partner HIV testing when HIVST is available at home. However, men testing positive through HIVST appeared less likely to link to care than men testing positive at a clinic. These results highlight the potential of HIVST in increasing HIV testing rates, while underscoring the importance of developing effective approaches to maximizing linkage to care among those testing positive through HIVST.