In a group of 13 female subjects submitted in three sessions to regularly alternating pink noise-steps, different techniques were proposed to cope with individual differences in cardiovascular responses. Application of a clustering algorithm to intercorrelation data of individual heart rates (HR) resulted in three subsets of 'sensitive' (S), 'moderately sensitive' (MS) and 'poorly sensitive' (PS) subjects. Only one pattern of HR changes under intermittent noise was observed, but clear-cut differences appeared between sensitivity subgroups which differed for amplitudes of HR responses to noise-steps and for adaptivity to repeated exposures. S subjects always reacted strongly (in fact, progressively strongly); PS subjects always reacted poorly; MS subjects reacted moderately but some of them with exposures either seemed to adapt to the noise or tolerated the noise less and less easily. Although exposure to intermittent noise for 35 min did not prevent a slowing trend of HR in all subjects, no sign of habituation of HR responses to noise-steps was present in any subgroup with repeated exposures. Mean levels of measured systolic and diastolic blood pressure decreased significantly with repeated exposures in MS and PS subjects, but not in S subjects.