This study evaluated the efficiency of a questionnaire as compared to a brief clinical examination for identification of clinically assessed craniomandibular disorder cases. 706 Danish children 14-16 yr old answered a self-administered questionnaire concerned with dysfunction symptoms and oral habits. In addition they were subjected to a clinical examination of signs of dysfunction from the masticatory system. Based on the present study a self-administered questionnaire for screening purposes used in an a priori healthy population of teenagers cannot be recommended. The answers to the questionnaire showed a low reproducibility and their ability to predict clinically assessed signs of dysfunction was inadequate. The most favorable combinations of single questions were calculated in order to optimize sensitivity, specificity, and predictive value of the questionnaire. A maximum of 65% of the clinically diseased subjects were identified by the questionnaire and only 41% of the test positive subjects had their subjective diagnoses confirmed at the clinical examination. The brief clinical screening procedure succeeded in identifying 92% of children with severely disturbed function. The subjects testing positively according to the brief clinical examination were children with either moderate or severe disturbances except for only one case. Based on a cost to benefit consideration a brief clinical examination is preferred to a self-administered questionnaire as a routine procedure for detection of craniomandibular disorders.