This study aimed to conduct a systematic review of research that has evaluated the outcomes of exercise-based interventions in anorexia nervosa (AN) patients. Method: We searched for studies in the MEDLINE, Embase, Web of Science, PsycInfo, Scopus and SPORTDiscus databases. We followed Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines throughout the selection process, and used Cochrane tools to assess the risk of bias in the primary studies. In cases where sufficient information was available, outcomes were pooled using meta-analytic techniques. Results: Ten studies were included in the review. Results showed similar changes in body mass index in both experimental and control conditions (difference in mean changes of −0.53, 95% CI −1.52 to 0.46). We found no differences in depression or anxiety as a result of implementing exercise programmes. Studies using resistance or strength exercise programmes reported an increase in lower and upper body strength and in skeletal muscle mass in the intervention groups. Conclusion: The incorporation of a supervised physical exercise programme to the usual treatment may be appropriate for AN patients. Exercise interventions should only be performed when they are safe for AN patients and should be focused on resistance/strength or flexibility exercises. The results from this systematic review should be interpreted with caution because the studies reviewed showed serious methodological problems.