Aims: The operational definitions for the Craniomandibular Index (CMI) were redesigned to conform precisely to those of the Research Diagnostic Criteria for Temporomandibular Disorders (RDC/TMD), resulting in a single examination protocol, the Temporomandibular Index (TMI). The objectives were to evaluate interexaminer reliability of the TMI as well as its criteria and construct validity for measurement of TMD severity. Methods: Interexaminer reliability of the TMI was assessed on 12 subjects. Criterion validity of the TMI was evaluated relative to the CMI, the latter having established validity. Construct validity of the TMI was evaluated for its capacity to differentiate TMD patients (n = 79) from normal subjects (n = 20) and to detect changes in severity over time. Results: The examiners' average TMI scores were 0.27 ± 0.19 (SD) and 0.26 ± 0.20. Agreement was excellent, with an intraclass correlation coefficient (ICC) of 0.93. The scores for the TMI and the CMI correlated highly, with an ICC of 0.97. Statistical contrasts between the symptomatic groups and the normal subjects were highly significant (P .001). In 20 TMD patients who underwent treatment for their disorder, their mean change of 0.12 from their pretreatment TMI scores was highly significant (P .001). Conclusion: This study has provided statistical evidence for the clinical reliability and validity of the TMI, which indicates that the RDC examination protocol is appropriate for determining TMD severity by the TMI algorithm, and diagnosis of TMD subtypes by the RDC algorithm. Parole chiave: temporomandibular disorder, reliability, validity, outcomes, Research Diagnostic Criteria, Craniomandibular Index