A scintillation camera connected to a dedicated computer system was employed for atraumatic CBF measurements in 43 patients after intravenous injection of 35 to 55 raCiluXe in saline solution. To validate this method results were compared to flow values in the same patients measured after intracarotid Xe injection. While the correlation between i.e. and i.v. values was not satisfactory when a high resolution collimator was used, a high sensitivity collimator improved count rates and yielded CBF values after i.v. in agreement with those from i.e. studies. For mean hemispheric flow, the correlation coefficient between the methods was 0.93 and the standard deviation of the i.v. value for a given i.e. value was 2.93. The correlation coefficients for 13 regions were between 0.55 and 0.85. These correlation coefficients are comparable to those obtained with multidetector equipment. Ischemic regions could be accurately detected, and the flow values in these areas were significantly related (r = 0.81). Values in 6 healthy volunteers were in the normal range reported by other investigators. One disadvantage of the camera is that measurements are restricted to one hemisphere, but selection of recording areas is not limited to the position of single probes and may be changed during analysis of the data permitting analysis of flow in irregularly-shaped, pathologically-perfused regions. The results indicate that the scintillation camera is a useful tool for clinical rCBF studies. Stroke, Vol 12, No 1, 1981