Effective implementation of evidence-based interventions in "real-world" settings can be challenging. Interventions based on externally valid trial findings can be even more difficult to apply in resource-limited settings, given marked differences-in provider experience, patient population, and health systems-between those settings and the typical clinical trial environment. Under the auspices of the Integrated Management of Physician-Delivered Alcohol Care for Tuberculosis Patients (IMPACT) study, a randomized, controlled effectiveness trial, and as an integrated component of tuberculosis treatment in Tomsk, Russia, we adapted two proven alcohol interventions to the delivery of care to 200 patients with alcohol use disorders. Tuberculosis providers performed screening for alcohol use disorders and also delivered naltrexone (with medical management) or a brief counseling intervention either independently or in combination as a seamless part of routine care. We report the innovations and challenges to intervention design, training, and delivery of both pharmacologic and behavioral alcohol interventions within programmatic tuberculosis treatment services. We also discuss the implications of these lessons learned within the context of meeting the challenge of providing evidence-based care in resource-limited settings.