Objective To compare effects of intra-articular injection of corticosteroid or viscoelastic agent alone or a combination of the two drugs for treatment of temporomandibular joint disorders. Methods A prospective quasirandomized controlled trial was conducted to compare 3 treatment protocols of intra-articular injections in the upper compartment of the joint immediately following arthrocentesis repeatedly every 10 days: ① triamcinolone 8 mg alone for 3 times, ② 1% hyaluronate 1 ml alone for 4 times, and ③ triamcinolone 8 mg for 2 times then 1% hyaluronate 1 ml for 2 times. Clinical examinations were done at baseline, 1, 12, and 24 months after end of the treatments. According to improvement of the symptoms and clinical signs, the effectiveness was graded in 3 classes: excellent, better, no change, or worse. The first two were classified as effective. Effective rates of the treatments and subgroups were compared statistically. Results Five hundred and sixty four patients with temporomandibular disorders were included and randomly allocated to 3 groups with 188 patients in each group. The rate of lost follow up at 24 months was 6.9% to 10.1%. At one month after treatment, the triamcinolone group had a better effective rate at 92.0%. At 12 and 24 months of follow up, effective rates of the hyaluronate group were better than those in the triamcinolone alone group (84.6% vs. 54.1% and 83.4% vs. 40.4%). Effective rates of combining triamcinolone and hyaluronate group at 12 and 24 months were 90.4% and 66.3% respectively. Conclusion Intra-articular injection of corticosteroid has better results in a short term and hyaluronate has better results in a long term in the treatment of temporomandibular joint disorders.