Arthroscopic surgery is an effective treatment for patients with temporomandibular disorders, releasing symptoms and restoring the mandibular function. In patients with poor arthroscopic outcomes, several options of treatment can be considered such as conservative nonsurgical therapy, open surgery, or a second arthroscopy. The purpose of this study was to evaluate our results after 619 arthroscopies. The clinical data of 619 arthroscopies performed between 1996 and 2015 were reviewed retrospectively. Outcome assessments were based on reductions in pain, measured using a visual analog scale (VAS), and improvement in maximal interincisal opening (MIO). The minimum follow-up period was 24 months. The incidence of TMJ reoperation in the 371 patients who underwent arthroscopic surgery was 5.9%, with a mean time between surgeries of 66.73 months. Significant improvement between presurgical and postsurgical pain and presurgical and postsurgical MIO at months 6 and 12 were evident. The mean of preoperative MIO was 30.84 mm, which increased to 35.92 mm 1-year postsurgery, these results being statistically significant (p < 0.05). In most of the cases the result of a new arthroscopy is satisfactory. Rearthroscopy of the TMJ is a valid and effective method for patients with a first unsuccessful arthroscopy.