The pattern of complaints of patients with the hyperventilation syndrome (HVS) was studied on the basis of the Nijmegen HVS Questionnaire (van Doorn, Colla, Folgering). This list was completed by 75 patients with the clinical diagnosis HVS. Non-metric principal components analysis (NMPCA) showed that the structure was three-dimensional, the dimensions being labelled: Shortness of breath (HVS-1), Peripheral tetany (HVS-II), Central tetany (HVS-III). The questionnaire's differentiating ability was investigated by comparing HVS patients with non-HVS persons (80 persons employed in health care). All three components had an unequivocally high ability to differentiate between HVS and non-HVS. Application of linear analysis of discriminance to HVS-I, HVS-II and HVS-III together yielded 93% correct classifications. Statistical double cross-validation resulted in 90 and 94% correct classifications. The sensitivity of the Nijmegen Questionnaire in relation to the clinical diagnosis was 91% and the specificity 95%. It is concluded that the questionnaire is suitable as a screening instrument for early detection of HVS, and also as an aid in diagnosis and therapy planning.