The most commonly used laboratory test for predicting malignant hyperthermia susceptibility is the caffeine halothane contracture test. However, the specificity and sensitivity of proposed North American diagnostic guidelines for this test have never been evaluated in a large, human study population. Therefore, the authors conducted a multiinstitutional, prospective study of skeletal muscle contracture responses in a subject population at low risk for malignant hyperthermia susceptibility to help determine the specificity of the proposed guidelines. Subjects were selected arbitrarily from a population of patients undergoing surgery unrelated to performance of a diagnostic muscle biopsy. Subjects were admitted to this study and were presumed nonsusceptible if there was no evidence of any of the following malignant hyperthermia risk factors: prior abnormal response to triggering anesthetic agents, myopathy, or family history of malignant hyperthermia susceptibility. The authors suggested rejection of the proposed diagnostic guidelines if an 85% specificity estimate among subjects could not be obtained. The authors analyzed the responses of 1,022 muscle fascicles, derived from 176 subjects, to the following: 1) separate administration of 3% halothane or incremental caffeine concentrations, or 2) the joint administration of 1% halothane and incremental caffeine concentrations. The following contracture results were obtained. First, for individual fascicles, 9.2% exceeded a greater than 0.7 g threshold for 3% halothane, 15.2% exceeded a greater than or equal to 0.2 g threshold for 2 mM caffeine, 32.4% exceeded a 1-g increase for less than 4 mM caffeine, 2.6% had a greater than 7% maximal increase in tension at 2 mM caffeine, and 63.5% had a "halothane caffeine-specific concentration" at less than or equal to 1 mM caffeine. Second, the percentages of subjects with 1 or more fascicles exceeding the proposed threshold were as follows: 45.8% for the four-component, 28.8% for the three-component, and 32.7% for the two-component contracture test. Third, the percentages of subjects with 1 or more fascicles exceeding the proposed threshold for both halothane and caffeine were as follows: 9.5% for 3% halothane and 2 mM caffeine, 2.0% for 3% halothane and 7% maximal increase in tension at 2 mM caffeine, and 11.0% for 1% halothane and 2 mM caffeine. Fourth, center-to-center differences were the major source of variation in the rate that subjects exceeded proposed thresholds. These data demonstrate that proposed diagnostic guidelines must be modified to improve specificity estimates before adoption by diagnostic centers.(ABSTRACT TRUNCATED AT 400 WORDS)