Physiologic factor analysis (PFA) has been applied to a set of dynamic positron emission tomography (PET)-generated images to extract fundamental, kinetic functions useful in compartmental modeling of PET data. The study was conducted to investigate PFA as a means of improving compartment models of tracer kinetics for the estimation of neuroreceptor binding characteristics from dynamic PET data. PFA derived factors avoid the problem of overlapping tissue types in compartmental estimates and also avoid errors in operator definition of regions of interest, since PFA is an automated method. Three factors were estimated: the first factor was identified as a sample of the free tracer in tissue compartment and accounted for a mean contribution of 41% to the total factor representation of the data; the second factor was identified as bound radioligand with a 33% mean contribution; and the third was identified as free tracer in blood with a 26% mean contribution. The PFA results obtained from the 14 human PET studies were compared to published results from animal studies using the same radioligand but where tissue samples were analyzed for radioactivity. The time-dependent behavior of compartmental activity in the two cases was similar.