Flow velocities (F.V.) in 65 patients admitted for subarachnoid hemorrhage (S.A.H.) were measured 4-7 times a week by Transcranial Doppler (T.C.D.). Patients were classified into 4 T.C.D. groups according to the highest mean flow velocity recorded in the M.C.A. during hospitalization: Group 1 (18 patients) with F.V. below 100 cm/s (normal), group 2 (19 patients) with F.V. between 100-150 cm/s (starting spasm), group 3 (23 patients) with F.V. between 150-200 cm/s (major spasm) and group 4 (5 patients) with F.V. over 200 cm/s (critical spasm). Based on clinical, radiological and ultrasound data as well as patient outcome (good results with no specific problems or with transient deficit, permanent deficit and pre- or postoperative death), the authors analyse the value of T.C.D. in the evaluation of vasospasms following S.A.H. The review involves 49 patients with surgically treated aneurysms (including two associated A.V.M.s), 9 patients who had suffered a S.A.H. of unknown origin and 7 patients who had died prior to surgery. T.C.D. is a non-invasive method of following post-S.A.H. spasms and the close correlation between the angiography and the T.C.D. makes pre-operative control angiographies unnecessary. A correlation between clinical status and T.C.D. was only observed in extreme cases where F.V. exceeded 200 cm/s or increased by 100 cm/s or more within 3 days (6 patients out of 65). M.C.A. spasm can, however, be underestimated by the T.C.D. approach in cases where there is an associated spasm of the infraclinoidal carotid artery. In these cases, classical Doppler evaluation of the cervical portion of the I.C.A. is indicated.(ABSTRACT TRUNCATED AT 250 WORDS)