Objectives: Expanding medication abortion to the pharmacy setting could help women access safe care. Studies have not evaluated pharmacy interventions. This non-inferiority study in Nepal investigated whether mifepristone-misoprostol medication abortion provided by trained nurses from pharmacies could be as effective and safe as when provided from government-certified abortion facilities. Methods: Nurses in two districts, Chitwan and Jhapa, were trained to provide medication abortion services from pharmacies. From October 2014-August 2015, we enrolled 605 women seeking abortion within 9-weeks’ gestation from 6 pharmacies (n=301) and 6 public facilities (n=304). Participants received medication abortion services, returning for a clinical exam and interview in 14-21 days. We used mixed effects models, accounting for clustering by site and provider, to compare complete abortion by study arm, and examined complications and satisfaction. Results: Among the 600 participants completing follow-up (>99%), the complete abortion rate was 98.0%, with effectiveness no lower among pharmacy participants (98.7%) than public facility participants (97.4%) (adjusted risk difference=1.5, 95% CI: -0.8, 3.8). Overall, 1.2% (1.7% pharmacy, 0.7% public facility) experienced a complication needing treatment, including infection and high fever. There were no serious adverse events. 95% of pharmacy participants and 96% of public facility participants reported they would prefer to return to the same setting for future services if needed. Conclusion: Mifepristone-misoprostol abortion can be as effective and safe when provided by trained nurses from pharmacy settings as from public health facilities. Findings support government expansion of medication abortion services through pharmacies by trained nurses to improve abortion access in Nepal.