OBJECTIVE::To determine whether patients have better outcomes when exercise instructions are provided using multimedia approaches compared with verbal or written instructions. DATA SOURCES::Electronic databases (MEDLINE, EMBASE, CINAHL, and PsychInfo) searched up to October 2018. STUDY SELECTION::Randomized controlled trials exploring exercise-based interventions for health conditions, and comparing instructions provided using multimedia approaches with conventional verbal or written instructions. RESULTS::Fourteen trials from seven countries were included, with a total of 2156 participants. Diagnoses included orthopaedic, neurological, pulmonary, cardiac, and women's health conditions. A meta-analysis of three trials (140 participants) provided very low-quality evidence that multimedia exercise instructions may be more effective than written instructions in improving exercise adherence (standardized mean difference (SMD) 0.60, 95% confidence interval (CI) -0.06 to 1.25). Two of nine trials that could not be included in the meta-analysis for adherence due to heterogeneity reported that multimedia exercise instructions were more effective than written instructions in improving exercise adherence. Four other meta-analyses (three trials each) found low- to high-quality evidence that provision of exercise instructions using multimedia is no more beneficial than paper-based instructions for patient-related outcomes of pain intensity (SMD 0.09, 95% CI -0.47 to 0.28); uptake of physical activity (SMD 0.07, 95% CI -0.08 to 0.23); or physical (SMD 0.21, -0.21 to 0.64) or emotional (SMD 0.16, 95% CI -0.04 to 0.36) domains of health-related quality of life. CONCLUSION::Multimedia approaches to exercise instruction may result in increased adherence compared with instructions provided in written or verbal format, but there is insufficient evidence to determine whether this results in improved patient outcomes.