Renal parenchymal transit time of the recently introduced radiopharmaceutical 99mTc-MAG3 (mercaptoacetylglycylglycylglycinel) was measured in 37 kidneys, using factor analysis to separate parenchymal activity from that in the collecting system. A new factor algorithm was employed, based on prior interpolative background substraction and use of the fact that the initial slope of the collecting system factor time-activity curve must be zero. The only operator intervention required was selection of a rectangular region enclosing the kidney (by identifying two points at opposite corners). Transit time was calculated from the factor time-activity curves both by deconvolution of the parenchymal factor curve and also by measuring the appearance time for collecting system activity from the collecting system factor curve. There was substantial agreement between the two methods. Factor analysis led to a narrower range of normal values than a conventional cortical region-of-interest method, presumably by decreasing crosstalk from the collecting system. In preliminary trials, the parenchymal transit time did not well separate four obstructed from seventeen unobstructed kidneys, but it successfully (P < 0.05) separated six transplanted kidneys with acute rejection or acute tubular necrosis from 10 normal transplants.