Clinical evidence indicates that parasympatholytic effects of tricyclic antidepressants increase with age. The aim of the present study was to determine the possible physiological reason for this phenomenon. Subjects included 23 patients (14 female) with major depression, melancholic type, and 23 age- and sex-matched healthy control subjects. Cardiac vagal tone was measured at rest using both spectral analysis and a time domain beat-to-beat method. Results of the spectral and time domain methods for the estimation of vagal tone used in this study were highly correlated in control subjects as well as in medicated depressed subjects. Both patients and control subjects showed an age-related decline in cardiac vagal tone. Tricyclic antidepressants decreased vagal tone significantly by 25-49% depending on age (20-60 years), although the age difference was not significant. The greater effect of tricyclic antidepressants on parasympathetic activity typically seen in older age groups may reflect the fact that predrug levels of vagal tone are already low in older patients. Measurement of vagal tone prior to drug administration may therefore be of prognostic value for anticholinergic side effects.