SUMMARY:This study assesses the quality, as a diagnostic test, of the main indicators of morphological sexual dimorphism through direct anthropometry, biostatistics tools and clinical epidemiology. This study used 284 skulls of adult Brazilians, of which 187 were male and 97 female. A study of the cross-evaluation of the diagnostic test was performed; it was a qualitative approach based on visual examination of 16 traditional indicators of morphological sexual dimorphism, where each indicator determined the level of accuracy, sensitivity, predictive values, likelihood ratios, and odds ratio. All indicators studied had high levels of accuracy (84.75-72.89%). The best indicators were found in traits whose formation is related to the insertion and action of major muscle groups. In 14 of the 16 indicators, intraobserver error was <10%. The best indicators of morphological sexual dimorphism were mastoid process, zygomatic bone, mandible, and roughness of the occipital bone. The authors concluded that morphological dimorphism indicators present an adequate performance as diagnostic tests, however, the values of accuracy and sensitivity must be matched with more robust indicators that are independent of the distribution of the sample, and integrate diagnostic errors such as the likelihood ratios, odds ratios, and positive predictive values.