Objectives: To study the diagnostic accuracy of the clinical examination of the Research Diagnostic Criteria (RDC) and of the dynamic and static pain tests for the recognition of temporomandibular disorder (TMD) pain and to improve the RDC accuracy by 1) changing the myofascial pain cutoff of 3 painful muscle palpation sites, or 2)omitting unreliable palpation sites. Methods: In 4 European dental faculties, a blind examination was performed in 125 chronic TMD pain patients, 88 chronic dental pain patients, and 121 pain-free subjects. Allocation was based upon the results of an oral history and a dental examination. Results: Sensitivity and specificity of the RDC were .88 and 45-.71, respectively. Increasing the myofascial pain cutoff better met the recommended levels for specificity and sensitivity of .70 and .90, respectively. When unreliable muscle palpation sites (i.e., the intraoral and submandibular sites) were omitted, the accuracy of the RDC/TMD examination did not change. For the dynamic and static pain tests, sensitivity (.65) and specificity (.84-.91) did not differ significantly from the recommended levels. Conclusion: A suspected TMD origin of orofacial pain is best confirmed by pain on the dynamic or static tests, while it is better denied by a negative outcome of the RDC examination. The intraoral and submandibular palpation sites of the RDC examination do not contribute to its diagnostic accuracy and can better be omitted, while the cutoff for a myofascial pain diagnosis should be increased.