Some participants will get HIV-infected in HIV prevention trials, despite risk reduction measures. The subsequent treatment responsibilities of sponsor-investigators have been widely debated, especially where access to antiretroviral therapy (ART) is not available. In this paper, we explore two accounts of beneficence to establish whether they can shed light on sponsor-investigator responsibilities. We find the notion of general beneficence helpful insofar as it clarifies that some beneficent actions will be obligatory where they can be dispensed without scuppering the trial. We (...) find the notion of specific beneficence helpful insofar as it directs investigators to attend to the needs of trial participants; however the range of interventions that could be provided remains unhelpfully broad. We then examine accounts of the investigator-participant relationship to narrow the range of interventions that investigators should provide, concluding that health-care, and HIV infection, are appropriate foci. We conclude that when investigators are able to meet the ART needs of their participants (e.g. referral, assisted referral or direct provision) without sacrificing trial quality, they must do so. However, there is little of this explicit direction to be found in the account of specific beneficence itself, but rather it is found in accounts of the relationship that are compatible with beneficence. (shrink)