The aim of this study was to investigate if a TMD-trained general dental practitioner could individually predict actual treatment outcome in selected patients diagnosed with temporomandibular disorders (TMD) with similar results as a TMD specialist. The patients were examined, individually predicted, treatment planned, treated and had their treatment outcome evaluated by the therapist, respectively. Out of 2618 patients referred to a TMD specialist clinic, 1086 patients started treatment. They were all divided into Muscle or Mainly TMJ symptoms. Prediction of the treatment outcome as Good or Dubious was based on the patient's history, the clinical and, sometimes, radiological findings. The degree of improvement was graded using a Numeric Rating Scale 0-100. A clinical important improvement, defined as an improvement of initial complaints of 50% or more, was judged as a correct prediction of Good treatment outcome. Seven-hundred-sixty-nine patients treated by the TMD specialist (Sample 1) was compared with 164 patients treated by the TMD-trained general dental practitioner (Sample 2). For patients with Muscle symptoms in Sample 1, a 50% improvement or more was reached by 93% of those predicted Good and 57% of those predicted Dubious. The corresponding figures in Sample 2 were 100% and 82%, respectively. In Sample 1, patients with Mainly TMJ symptoms reached a 50% improvement or more in 94% of those with prediction Good and 73% of those predicted Dubious. In Sample 2 the figures were 100% and 87%, respectively. ATMD-trained general dental practitioner could individually predict treatment outcome with similar results as a TMD specialist in selected patients diagnosed with TMD. Whether the method is possible to generalize has to be investigated further.