The particular questions asked in our study were: 1. does the individual reproducibility of the cardiovascular reflex tests differ between healthy controls and patients suffering from type I diabetes mellitus and 2. if there is a difference, do the different cardiovascular reflexes vary in this regard? Nine healthy controls (4 women, 5 men, age 31±2.1 years) and 11 type I diabetics (4 women, 7 men, age 30.9±5.6 years, duration of diabetes 3–23 years) underwent the following tests 6 times in a 12-h period (07⁰⁰ to 19⁰⁰): variation of heart rate during deep breathing (E/I ratio). variation of heart rate during lying and standing (tachycardia/bradycardia or 30/15 ratio), Valsalva maneuver (Valsalva ratio), response of diastolic blood pressure to sustained hand grip, and response of systolic blood pressure to posture. The test results did not indicate a diurnal fluctuation nor were they systematically influenced by antecedent insulin injections or meals, either in diabetic patients or in healthy, controls. The 11 diabetics had significantly lower intraindividual variations of E/I and Valsalva ratios than the controls (p<0.05, p<0.001, respectively). In the diabetics with parasympathetic failure the intraindividual variabilities of all cardiovascular reflex responses were lower than those of the patients with an intact autonomic nervous system as well as those of the control subjects. On the contrary, in the diabetic patients without autonomic neuropathy, only the intraindividual variability of the Valsalva maneuver was significantly attenuated (p<0.025), compared with the healthy volunteers. To conclude, the more pathological the single test result, the greater is its reproducibility and its clinical significance.