It is generally agreed that Langham's model for urinary excretion of Pu substantially overestimates the systemic burden several years after exposure. Improved estimates can be derived from information obtained since the development of that model, including comparative urine and autopsy data for occupationally exposed persons; reanalyzed and updated data for human subjects injected with Pu; and a large body of general physiological and Pu-specific information on the processes governing the behavior of Pu in the body. We examine modeling approaches based on each of these sets of information and show that the three approaches yield fairly consistent estimates of the urinary excretion rate over three decades after contamination of blood. Estimates from the various approaches are unified to obtain a single set of predicted urinary excretion rates that, in effect, is based on all three bodies of information. A simple method is described for using these excretion rates to estimate intakes and systemic burdens of Pu.