A cross-sectional survey of the prevalence of asthma and bronchial hyperreactivity among schoolchildren (7 to 11 years old) was carried out in three areas of the Latium region (Central Italy). Out of 1,777 children tested with methacholine challenge (MCT), 15.1 percent had a 20 percent fall in FEV1 after a provocative concentration (PC20FEV1) of 4 mg/ml of methacholine or less; 69.7 percent had a PC20FEV1 less than 64.0 mg/ml, whereas 50.3 percent were nonresponders. Two continuous measures of bronchial responsiveness, the slope (percentage of change in FEV1 per mg/ml of methacholine) and the area under the dose response curve, were calculated in order to avoid the loss of information in nonresponders. Applying a receiver operating characteristic (ROC) curve analysis, the three estimators did not show any statistically significant difference in their overall performance in detecting asthma (ROC areas: PC20FEV1 = 0.683, slope = 0.681, area = 0.702 or asthma-like symptoms. The log transformation of slope, having a unimodal and slightly skewed shape, is an appealing continuous measure of bronchial responsiveness useful for epidemiologic studies. The final choice of an appropriate estimator of the concentration-response curve to methacholine, however, depends upon both the statistical tests or the modelling procedures to be used and clarification of the prognostic value of different indices of bronchial responsiveness.