Reflecting an increasing awareness of the importance of treatment adherence on outcomes in psychiatric populations, the National Institute of Mental Health (NIMH) convened a panel of treatment adherence researchers on September 27-28, 2007 to discuss and articulate potential solutions for dealing with methodological adherence research challenges. Panel discussions and presentations were augmented with targeted review of the literature on specific topics, with a focus on adherence to medication treatments in adults with serious mental illness. The group discussed three primary methodological areas: participants, measures, and interventions. When selecting patients for adherence-enhancing interventions (AEIs), a three-tier model was proposed that draws from the universal (targeting all patients receiving medication treatment for a specific condition, regardless of current adherence), selective (targeting patients at risk for nonadherence), and indicated (targeting patients who are currently nonadherent) prevention model and emphasizes careful patient characterization in relevant domains and appropriate matching of interventions to the selected population. Proposals were also made to reduce problematic selection biases in patient recruitment and retention. The panel addressed the pros and cons of various methods that can be used to measure adherence, and concluded that it is appropriate to use multiple measures whenever possible. Finally, the panel identified a broad range of intervention approaches, and conditions under which these interventions are likely to be most effective at reducing barriers to adherence and reinforcing adherence behavior.