Although patient attrition might be a serious threat to the validity of treatment-outcome studies on temporomandibular disorders (TMD), studies on TMD patient attrition are scarce. Of the 1405 consecutive TMD patients examined in a recent 10-year period, 367 (26.1%) drop-out pati ents or patients identified with a control group were sampled. A mailed questionnaire failed to reach 41 patients, and 203 (62.3%) were returned. The questionnaire elicited information on reasons for dropping out, changes in symptoms, treatment received in other clinics after dropping out, present treatment needs, and current signs and symptoms. Dropouts were divided into two groups: (1) those who failed to show up for their first scheduled appointment after the clinical examination; (2) those who failed to complete treatment. A group of patients who were judged by the examiner not to need treatment were included as a control group. The main reasons for dropping were environmental obstacles, perceived improvement of the disease, and dissatisfaction with services. Only 21.7% considered themselves to be in need of treatment, and only 10.3% had visited other clinics after dropping out. Only 8.9% complained of the continued aggravation of symptoms, whereas 57.6% reported improvement. In addition, pain, dysfunction, and daily activity limitation tended to improve with time, although temporomandibular joint noise tended to persist. These results suggest that TMD signs and symptoms tend to devrease in patients after dropping out, and that the natural fluctuation of TMD signs and symptoms should be taken into consideration when treating TM