METHODS Medline, Cochrane, Web of Science, SciELO and Embase databases from 1993-2014. METHODS Randomised Controlled Trials (RCTs) comparing a manual therapy physical therapy intervention to a reference group (placebo intervention, controlled comparison intervention, standard treatment or other treatment). METHODS Two independent reviewers abstracted data and assessed quality and clinical relevance of each paper. In case of disagreement, a third reviewer was consulted. The PEDro (Physiotherapy Evidence-Based Database) scale was used to assess the methodological quality of the studies. RESULTS Eight studies were included. The number of patients in the studies ranged from 30-93. Seven out of the eight studies presented high methodological quality. Treatment effect size was calculated for pain, maximum mouth opening (MMO) and pressure pain threshold (PPT). There was moderate and low evidence that myofascial release and massage techniques are more effective than placebo or no intervention for MMO and pain outcomes respectively. There was also moderate evidence that no significant difference exists between myofascial release and toxin botulinum for improvement on the same outcomes. There was low to high quality evidence that upper cervical spine thrust manipulation or mobilisation techniques are more effective than control, while thoracic manipulations are not. Overall there was moderate-to-high evidence that MT techniques protocols are effective. Methodological heterogeneity of the trial protocols frequently contributed to a decrease in the quality of evidence. CONCLUSIONS There is widely varying evidence that MT improves pain, MMO and PPT in subjects with TMD signs and symptoms, depending on the technique. Further studies should consider using standardised evaluations and better study designs to strengthen clinical relevance.