Researchers in medical education have been placing increased emphasis on "outcomes research," or the observable impact of educational interventions on patient care. However, although patient outcomes are obviously important, they should not be the sole focus of attention in medical education research. The purpose of this perspective is both to highlight the limitations of outcomes research in medical education and to offer suggestions to facilitate a proper balance between learner-centered and patient-centered assessments. The authors cite five challenges to research using patient outcomes in medical education, namely (1) dilution (the progressively attenuated impact of education as filtered through other health care providers and systems), (2) inadequate sample size, (3) failure to establish a causal link, (4) potentially biased outcome selection, and (5) teaching to the test. Additionally, nonpatient outcomes continue to hold value, particularly in theory-building research and in the evaluation of program implementation. To educators selecting outcomes and instruments in medical education research, the authors offer suggestions including to clarify the study objective and conceptual framework before selecting outcomes, and to consider the development and use of behavioral and other intermediary outcomes. Deliberately weighing the available options will facilitate informed choices during the design of research that, in turn, informs the art and science of medical education.