The maximum value of capillary filtration coefficient (CFC) in maximally vasodilated cat skeletal muscle is disputed. It was hypothesized that the wide range of reported values was caused by the inability of gravimetric and volumetric measurements of tissue volume to separate transcapillary filtration from vascular volume changes. Consequently, we developed a method of measuring filtration rates from changes in venous protein concentration using Evan's blue-labeled albumin in the isolated hindlimb (pentobarbital sodium anesthesia). The filtration coefficient (PFFC) calculated from these filtration rates after a step in venous pressure should not be influenced by vascular volume changes. When the perfusate flow rate through the hindlimb was greater than 15 ml.min-1.100 g muscle-1, PFFC was 0.0085 +/- 0.0015 (SD, n = 8) ml.min-1.mmHg-1.100 g muscle-1. PFFC was observed to be unvarying from 1 to 12 min after the venous pressure elevation, in contrast to CFC values, which fall during the same period. It is argued that the difference between CFC and PFFC values is caused by vascular volume changes.