CONTEXT: The availability of trained abortion providers is limited in India. Allowing ayurvedic physicians and nurses to perform medication abortions may improve women's access to the procedure, but it is unclear whether these clinicians can provide these services safely and effectively. METHODS: Allopathic physicians, ayurvedic physicians and nurses (10 of each), none of whom had experience in abortion provision, were trained to perform medication abortions. In 2008-2010, these providers performed medication abortions in five clinics in Bihar and Jharkhand for 1,225 women with a pregnancy of up to eight weeks' gestation. A two-sided equivalence design was used to test whether providers' assessments of client eligibility and completeness of abortion matched those of an experienced physician "verifier," and whether medication abortions performed by nurses and ayurvedic physicians were as safe and effective as those done by allopathic physicians. RESULTS: Failure rates were low (5-6%), and those for nurses and ayurvedic physicians were statistically equivalent to those for allopathic physicians. Provider assessments of client eligibility and completeness of abortion differed from those of the verifier in only a small proportion of cases (3-4% for eligibility and 4-5% for completeness); these proportions, and rates of loss to follow-up, were statistically equivalent among provider types. No serious complications were observed, and services by all three groups of providers were acceptable to women. CONCLUSION: Findings support amending existing laws to improve women's access to medication abortion by expanding the provider base to include ayurvedic physicians and nurses.