Specificity and sensitivity of two physiological markers for hot flushes were investigated. One marker, proposed by Freedman, is an increase of sternal skin conductance, the second marker, proposed by Swartzman, is a physiological profile which consists of skin conductance changes in combination with circulation changes. In our laboratory 20 menopausal women, 15 with frequent hot flushes and 5 without hot flushes, and 5 women with regular menstrual cycles were continuously monitored for 2.5 h on subjective hot flush experience, sternal and palmar skin conductance, dorsal and palmar finger temperature and pulse blood volume. Increase in sternal skin conductance proved to be very specific in contrast to Swartzman's physiological profile, although it was less sensitive. Receiver operating characteristics revealed that an increase combined with a preceding decrease in sternal skin conductance as most specific for, and most sensitive to, subjectively reported hot flushes. This was confirmed by a cross-validation with 34 'flushing' menopausal women.