Several theoretical and practical aspects of regional cerebral blood flow measurements using stable xenon gas and CT are discussed. It is shown that by comparing the enhancement at any time T1 with that at saturation or any other time T2, the need to use arbitrary means to bring the arterial concentration data and the CT enhancement data to the same system of measurement units can be eliminated. If CT is performed continuously during the washin phase, say at intervals of 1 min, least squares analysis of the enhancement data can be used to obtain the best possible estimates for the flow rate constant kappa and the saturation enhancement. However, if only a limited number of scans can be performed, as may be the case in human studies, it is also possible to get a good estimate of kappa from a knowledge of the ratio of the enhancement at any time T1 with that at any other time T2. Combinations of T1 = 2.0 min and T2 = 4.0 min, T1 = 1.0 min and T2 = 6.0 min, or T1 = 2.0 min and T2 = 5.0 min were found to be the most convenient. It is also shown that the end-tidal xenon concentration in the exhaled air can be accurately assessed indirectly by measuring the oxygen, CO2, and water vapor concentrations, thereby eliminating the need for more expensive methods involving the use of a mass spectrometer or a thermal conductivity gas analyzer.