PredictionModels I this issue, the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) investigators propose an annotated checklist for transparent reporting of prediction or prognostic models (1). The accompanying 22 000-word “Explanation and Elaboration” (2), with more than 500 references from statistics, epidemiology, and clinical decision making as well as from the applied clinical literature, should serve as an important resource for model developers. Applications span prediction (for diagnosis from cross-sectional data) and prognosis (for outcomes after longitudinal follow-up) across many clinical settings. These TRIPOD documents represent serious efforts to synthesize best practices for authors and readers. Because the TRIPOD checklist seeks primarily to improve reporting, the immediate benefit should be more comprehensive expositions of model development and validation. Dissemination is broad: The checklist is being published simultaneously in 10 journals that span the disciplines of general medicine and clinical practice, obstetrics and gynecology, urology, surgery, cardiology, diabetes, cancer, clinical biomedical science, and epidemiology. Longer-term benefits might flow from TRIPOD's success in leading the research and clinical communities to ask the right questions and consider the challenges in developing, validating, and using prediction models. As TRIPOD suggests, although there is no single correct method for modeling, authors should be able to justify their assumptions and describe their approaches. TRIPOD's comprehensive outline of modeling options should encourage informed choices and transparent presentations to further more general goals of reproducible research (3). The current version of TRIPOD is not the final word on prediction model development and validation. Methodological advances and critiques appear almost monthly and can quickly render guidelines obsolete. Fortunately, TRIPOD will maintain a Web site for updates. Complete reporting alone cannot generate clinically useful models. In our review of TRIPOD, we found the following areas of continuing concern for model developers and users.