We investigated the variability of absolute blood flow velocity (BFV) and task-induced BFV change over consecutive cognitive tasks and compared two methods of baseline determination that are used to calculate relative BFV changes. Bilateral transcranial Doppler ultrasonography recordings of BFV in the middle cerebral arteries was performed in 90 right-handed volunteers during 13 cognitive tasks and their preceding rest periods. Both absolute BFV and elicited BFV changes between rest and activation significantly decreased over successive tasks. Instead of calculating an averaged baseline value, our results suggest that the rest phase immediately preceding the activation phase should be selected for baseline measurement. The between-task habituation effect could be due to a fading of attentional resources during the sustained and demanding performance.