PURPOSE: The purpose of this systematic review was to summarise the evidence from randomised controlled trials examining the effectiveness of physical therapy compared with care as usual or a wait-list condition on eating pathology and on physiological and psychological parameters in patients with anorexia and bulimia nervosa. METHOD: EMBASE, PsycINFO, PubMed, Cumulative Index to Nursing and Allied Health Literature, Physiotherapy Evidence Database and The Cochrane Library were searched from their inception until February, 2013. Articles were eligible if they utilised a randomised controlled trial design, compared physical therapy with a placebo condition, control intervention, or standard care and included patients with anorexia and bulimia nervosa. The methodological quality was assessed with the Jadad scale. RESULTS: Eight randomised controlled trials involving 213 patients (age range: 16-36 years) met all selection criteria. Three of the 8 included studies were of strong methodological quality (Jadad score>/=3). Major methodological weaknesses were attrition and selection bias. The main results demonstrate that aerobic and resistance training result in significantly increased muscle strength, body mass index and body fat percentage in anorexia patients. In addition, aerobic exercise, yoga, massage and basic body awareness therapy significantly lowered scores of eating pathology and depressive symptoms in both anorexia and bulimia nervosa patients. No adverse effects were reported. CONCLUSIONS: The paucity and heterogeneity of available studies limits overall conclusions and highlights the need for further research. Implications for Rehabilitation Supervised physical therapy might increase weight in anorexia nervosa patients. Aerobic exercise, massage, basic body awareness therapy and yoga might reduce eating pathology in patients with anorexia and bulimia nervosa. Aerobic exercise, yoga and basic body awareness therapy might improve mental and physical quality of life in patients with an eating disorder.