Two methods are described for calculation of regional cerebral blood flow from completed tomographic data of radioactive inert gas distribution in a slice of brain tissue. It is assumed that the tomographic picture gives the average inert gas concentration in each pixel over data collection periods of 30 to 60 sec. In the early picture method a single picture taken during maximum inert gas concentration (by intraarterial injection or by inhalation) is analyzed. In the sequence of pictures method the alterations in local inert gas concentration are followed over time. The proposed methods are implemented using synthetic data of xenon-133 emission computed tomography and some of the difficulties likely to be encountered in practice are stressed.