BACKGROUND: Replication studies evaluate technologies in usual use settings. METHODS: We conducted a clinical trial to determine whether reductions in clinical and economic results observed in a previous study could be replicated in a larger setting. Subjects were randomized to receive intervention (email notifications for sentinel health events sent to their care managers) or control. MAIN OUTCOME MEASURES: The primary outcome was the rate of emergency department visits for low severity conditions. Secondary outcomes included: medical costs and other clinical event rates. RESULTS: We randomized 13,454 individuals (intervention, 6740; control, 6714). Subjects in both groups had similar rates of clinical events and medical costs. CONCLUSION: The use of email notifications to care managers was associated with no reductions in clinical events or medical costs.