Courses in Evidence-Based Medicine (EBM) for doctors have consistently demonstrated short-term improvements in knowledge. However, there is no strong evidence linking EBM training to changes in clinical practice or patient outcomes. This systematic review investigates whether EBM training leads to sustained improvements in doctors’ knowledge and practice behaviors that may also facilitate changes in patient outcomes and experiences. A literature search was undertaken in Ovid Medline, Ovid Embase, The Cochrane Library, ERIC and Scopus. Studies published from 1997 to 2016 that assessed outcomes of EBM educational interventions amongst doctors and used measures of knowledge, skills, attitudes, practice or patient outcomes were included. Fifteen studies were included in the analysis: four randomized controlled trials (RCTs), three non-RCTs, and eight before-after (longitudinal cohort) studies. Heterogeneity among studies prevented meaningful comparisons. Varying degrees of bias due to the use of subjective measures were identified, limiting study validity. Results showed that EBM interventions can improve short-term knowledge and skills, but there is little reliable evidence of changes in long-term knowledge, attitudes, and clinical practice. No study measured improvement in patient outcomes or experiences. EBM training for medical practitioners needs to incorporate measures of behavioral changes while incorporating patient outcomes and experience measures.