The majority of the maternal and perinatal deaths are preventable through improved emergency obstetric and newborn care at facilities. However, the quality of such care in India has significant gaps in terms of provider skills and in their preparedness to handle emergencies. We tested the feasibility, acceptability, and effectiveness of a "skills and drills" intervention, implemented between July 2013 and September 2014, to improve emergency obstetric and newborn care in the state of Karnataka, India. Emergency drills through role play, conducted every 2 months, combined with supportive supervision and a 2-day skills refresher session were delivered across 4 sub-district, secondary-level government facilities by an external team of obstetric and pediatric specialists and nurses. We evaluated the intervention through a quasi-experimental design with 4 intervention and 4 comparison facilities, using delivery case sheet reviews, pre- and post-knowledge tests among providers, objective structured clinical examinations (OSCEs), and qualitative in-depth interviews. Primary outcomes consisted of improved diagnosis and management of selected maternal and newborn complications (postpartum hemorrhage, pregnancy-induced hypertension, and birth asphyxia). Secondary outcomes included knowledge and skill levels of providers and acceptability and feasibility of the intervention. Knowledge scores among providers improved significantly in the intervention facilities; in obstetrics, average scores between the pre- and post-test increased from 49% to 57% (P=.006) and in newborn care, scores increased from 48% to 56% (P=.03). Knowledge scores in the comparison facilities were similar but did not improve significantly over time. Skill levels were significantly higher among providers in intervention facilities than comparison facilities (mean objective structured clinical examination scores for obstetric skills: 55% vs. 46%, respectively; for newborn skills: 58% vs. 48%, respectively; P<.001 for both obstetric and newborn), along with their confidence in managing complications. However, this did not result in significant differences in correct diagnosis and management of complications between intervention and comparison facilities. Shortage of trained nurses and doctors along with unavailability of a consistent supply chain was cited by most providers as major health systems barriers affecting provision of care. Improvements in knowledge, skills, and confidence levels of providers as a result of the skills and drills intervention was not sufficient to translate into improved diagnosis and management of maternal and newborn complications. System-level changes including adequate in-service training may also be necessary to improve maternal and newborn outcomes.