A retrospective study was performed to investigate the short-term therapeutic efficacy of non-steroidal anti-inflammatory drugs (NSAIDs) on temporomandibular disorders (TMD) with pain. TMD patients with TMJ pain and/or muscle pain were treated with amfenac sodium hydrate (150 mg/day) or loxoprofen sodium hydrate (180 mg/day), and the effects on TMJ pain, muscle pain, and maximum mouth opening range were examined after 7 days of administration. All subjects were classified in accordance with the subtype classification of TMD determined by the Japanese society for the temporomandibular joint following MRI examination. The efficacy rate for the pain was 85.7% in Type I, 87.0% in type II, 68.8% in type IIIa, 52.6% in type IIIb, 75.0% in type IV, respectively. There was no connection between the therapeutic efficacy for the pain and the patient's age. A significant difference of effectiveness was observed among type I, II and IIIb (p < 0.05). As for muscle pain, the therapeutic efficacy rate was 85.7% in no disc displacement group and 50.0 % in disc displacement group, respectively. The effectiveness differed significantly between the no disc displacement group and the disc displacement group (p < 0.05). As for TMJ pain, the therapeutic efficacy rate was 87% in the no disc displacement group, 72.0% in the disc displacement with reduction, 60.0% in the disc displacement without reduction group, respectively, effectiveness was noted to be significantly different between the no disc displacement group and the disc displacement without reduction group (p < 0.05). In patients with trismus (maximum mouth opening range < 40mm), no significant increase of the maximum mouth-opening range was observed after administration of NSAIDs.