Little information is available on the methodological and physiological limits of repeatedly testing changes of blood pressure (BP) and heart rate (HR) during autonomic testing. The retest variation determines the value of follow-up investigations for the assessment of the natural course of patients with autonomic failure or the efficacy of treatment. 34 healthy volunteers between 18 and 85 years of age were tested twice within a mean period of one year (75-637 days). BP and HR were measured continuously with a Finapress device during active change of posture, passive tilt, deep breathing, Valsalva manoeuvre, isometric muscle exercise and mental arithmetic. The differences to baseline resting values before each manoeuvre were analysed. Although differences between the two occasions were comparatively small with mostly 1-7 mmHg for BP and 0.3-4 bpm for HR, the high standard deviations did not allow the calculation of useful normal ranges, as they would exceed meaningful (patho-) physiological changes over time. Therefore, a single repetition of a series of standard autonomic function tests did not appear to be suitable for follow-up observations. The calculation of the components of the coefficients of variation showed a predominance of the within variance as the cause of the poor reproducibility for the BP measurements, while changes in HR were more influenced by the between variation. The combination of the selection of some tests with a high between variance and the reduction of the within variance by performing 4, 5 retest runs can increase the reliability of standard autonomic function tests regarding follow-up assessments.