The Cramer-Rao inequality and the Fisher information matrix have been used together with a Monte Carlo computer simulation to develop graphs for the determination of the peak count rate required to obtain radionuclide clearance parameters with a specified precision. Analytical relationships are derived for the coefficient of variation of each parameter estimate as a function of the clearance rate (A), the peak count rate (A), the scintillation camera sampling interval, and the number of sampling intervals. For a constant data analysis time, the coefficient of variation of each parameter is shown to be nearly independent of the number of data points if the number of frames exceeds 5. Thus, given estimates of the expected clearance rate and the data collection time, an investigator may determine the peak count rate required to measure the clearance parameters A and A with coefficients of variation less than a specified value. For a given peak count rate the results show that the clearance rate coefficient of variation for a bolus injection will always be smaller than for the washin to equilibrium procedure. An algorithm is presented for generating a regional xenon-133 lung clearance rate distribution with constant coefficient of variation and applied to experimental airway bolus response data in the dog.