The timing for surgery on ruptured intracranial aneurysms remains a difficult question and the choice of the day of operation depends greatly from the occurrence of a vasospasm. On a previous paper, the value of the cerebral blood flow (CBF) measurement by intravenous injection of Xenon 133 was demonstrated to be efficient for the prediction of vasospasm only when done between day 4 and day 8 after bleeding. Moreover the efficiency of the measurement was much greater if the evolution of the CBF values between D0 and D8 was considered, but this method was incompatible with early surgery. It suggested the interest of a dynamic study of the CBF by the same method. On a series of 43 patients, the comparison between basal CBF values and reactivity of CBF values to intravenous injection of 1 gram of acetazolamide for the prediction of clinical vasospasm is presented. The series comprises 32 early admitted patients (74%) and 31 operated patients (16 of them between D0 and D3 after bleeding, 15 others after D4). The efficiency of the CBF reactivity study for the prediction of the clinical vasospasm appears much greater than that of the basal CBF value, even during the first three days after bleeding, but not for the prediction of the final clinical outcome. The method is immediate and compatible with early surgery. What precisely is evaluated by this method on the physiopathology of the vasospasm remains disputable, but the operability of the patients seems to be explored.