BACKGROUND: Handoff miscommunications are a leading source of medical errors. Harmful medical errors decreased in pediatric academic hospitals following implementation of the I-PASS handoff improvement program. However, implementation across specialties has not been assessed. OBJECTIVE: To determine if I-PASS implementation across diverse settings would be associated with improvements in patient safety and communication. DESIGN: Prospective Type 2 Hybrid effectiveness implementation study. SETTINGS AND PARTICIPANTS: Residents from diverse specialties across 32 hospitals (12 community, 20 academic). INTERVENTION: External teams provided longitudinal coaching over 18 months to facilitate implementation of an enhanced I-PASS program and monthly metric reviews. MAIN OUTCOME AND MEASURES: Systematic surveillance surveys assessed rates of resident-reported adverse events. Validated direct observation tools measured verbal and written handoff quality. RESULTS: 2735 resident physicians and 760 faculty champions from multiple specialties (16 internal medicine, 13 pediatric, 3 other) participated. 1942 error surveillance reports were collected. Major and minor handoff-related reported adverse events decreased 47% following implementation, from 1.7 to 0.9 major events/person-year (p < .05) and 17.5 to 9.3 minor events/person-year (p < .001). Implementation was associated with increased inclusion of all five key handoff data elements in verbal (20% vs. 66%, p < .001, n = 4812) and written (10% vs. 74%, p < .001, n = 1787) handoffs, as well as increased frequency of handoffs with high quality verbal (39% vs. 81% p < .001) and written (29% vs. 78%, p < .001) patient summaries, verbal (29% vs. 78%, p < .001) and written (24% vs. 73%, p < .001) contingency plans, and verbal receiver syntheses (31% vs. 83%, p < .001). Improvement was similar across provider types (adult vs. pediatric) and settings (community vs. academic).