In the campaign to implement evidence-based practice, the current single-hierarchy model of levels of evidence fails to incorporate at parity all types of research evidence that are valuable in the practice of occupational therapy. A new model, originally developed by Borgetto et al. (2007) and modified and expanded, is presented. By separating the evidence-level criteria of internal and external validity, by incorporating explicitly the evidence provided by qualitative studies, and by retaining the critical notion of rigor, a pyramidal evidence model emerges. This model, the Research Pyramid, aligns itself with the revised model of evidence-based medicine and, more important, with the basic modes of clinical reasoning in occupational therapy. It constitutes a beginning attempt to order evidence-based practice in accordance with the epistemology of the profession. It may better guide occupational therapy research and meta-synthesis and their incorporation into practice decisions.