We consider the management of safety signals that arise in the context of endpoint trials. Such trials have design features which permit the detection of signals that otherwise may be missed in more traditional settings. However, the approaches that are typically employed to address them are less than satisfactory in terms of both the analytical techniques and the communication of the findings. In this paper, alternative approaches are evaluated, and recommendations are given for a comprehensive strategy that involves appropriate use of study data as well as information external to the study. It is emphasized that formulation and implementation of an effective communication plan that enables informed decision making should be an integral component of the strategy.