Background: Systematic reviewers disagree about the ability of observational studies (OS) to answer questions about the benefits or intended effects of interventions and the inclusion of these study designs in reviews. Objectives: To present and discuss the basic principles for selection and inclusion of OS to assess benefits in systematic reviews of comparative effectiveness across clinical treatment interventions. Methods: A conceptual model and recommendations were developed using a consensus process by members of the methods work group of the Effective Health Care Program of the US Agency for Health Care Research and Quality. Results: In considering whether to use OS in comparative effectiveness reviews for addressing beneficial effects, reviewers should answer two questions: 1) Is the trial evidence sufficient to address all aspects of the review question, including the population, specific intervention, comparator, outcomes, and setting? In other words, are the available trial data applicable to the needs of the user (which are reflected in the review questions)? 2) Will OS provide valid and useful information? The latter question involves: a) refocusing the review questions according to the gaps identified in the trial evidence; b) assessing whether OS can address these gap’ questions; and c) assessing how potential biases, magnitude of benefit, heterogeneity of effects, and random variation may influence the results of OS. A conceptual model encompasses these questions and processes. Conclusions: Because it is unusual to find sufficient evidence from trials to answer all key questions in a systematic review of comparative effectiveness, the inclusion of OS for questions of benefit in comparative effectiveness reviews should always be considered. Reviewers should explicitly state a priori their decision rules to include or exclude OS and fully describe the rationale for that decision based on gaps in the trial literature and on the potential validity of OS to address those questions.