The most widely used index to evaluate the Cerebral Autoregulation System is the autoregulatory index ARI proposed by Aaslid and Tiecks. Although it is often used in clinical research and is also included in some commercial equipment it has a major drawback: it exhibits great variability even when used in the same patient. It also produces many false positives. In this work an alternative measurement to the Aaslid-Tiecks ARI directly taken from the thigh cuffs maneuver signals and inspired in the Ziegler-Nichols reaction curve method is presented. This method uses only two parameters to characterize a response to the thigh cuffs maneuver. Once the values of these parameters are obtained, a correspondence between them and the parameters of the Aaslid-Tiecks, ARI, is computed using linear regression. Results from 16 healthy subjects to which 6 different thigh cuffs maneuvers were applied showed that the proposed direct measurement method reduces the variability of the ARI from 30% ± 21% to 9,7% ± 4,2% and also significantly reducing the number of false positives. © Springer-Verlag Berlin Heidelberg 2007.