The distribution of pulmonary blood flow and ventilation within the lung may be measured with ¹³³Xe. The recorded activity over the lung is a function of lung shape, spatial sensitivity of the detector, scattered radiation, and, if scanning, the time constant and scanning speed of the recording apparatus. By a study of isolated lung models a method was developed to overcome the problems of scattered radiation and spatial sensitivity in the determination of the no-flow point, and it was found that for accurate recording the product of the time constant (sec) and scanning speed (cm/sec) should be less than 0·2. The level at which flow ceases is 1·3 (±1·7) cm above the point at which alveolar pressure equals arterial pressure.In human subjects it is impossible to measure accurately the no-flow point due to the large detector field, scattered radiation and residual radiation in extra-thoracic tissue. Despite the fast scanning speed of 4 cm/sec for human subjects, it is suggested that an estimate of the mean linear rate of decrease in radioactivity in the lung could be obtained with a time constant of 0·2 sec.