Background: Nonemergent communications between nursing staff and residents often occur randomly throughout the workday at teaching institutions, resulting in inefficiency and interruptions of scheduled resident educational conferences or periods of overnight rest. Objective: The aim of this study was to determine whether a quality improvement communication strategy can reduce interruptive notifications during scheduled resident didactic sessions. Methods: We conducted a prospective, mixed-methods trial of a communication strategy between nurses and residents in 3 neurology units of an academic institution. We measured content and volume of notifications from nurses to residents during scheduled conferences and overnight and categorized these notifications according to levels of urgency. We implemented an intervention requiring multidisciplinary evening work rounds between the on-call resident and charge nurse and batching of nonemergent communications. Survey data were collected. Results: The total census during the 6-month study period was 728 patients. Of the 704 resident-reported nursing notifications, 50% could be safely deferred until change of shift. After the intervention, there was a 52% decrement of notifications (0.52 vs 0.25 notifications per patient, P < .01). The preintervention survey response rate was 65% for nurses (46/71) and 100% for residents (26/26), whereas postintervention response rates were 51% for nurses (36/71) and 96% for residents (24/25). Both nurses and residents responded more favorable in the postintervention period across multiple items that addressed patient safety, utility, and satisfaction (P < .01 for comparisons). Conclusions: A scheduled nurse-resident communication system reduced interruptive calls during educational conferences and overnight by half. This intervention was perceived as safe and effective and resulted in higher satisfaction among participants.