3H-Androgens in rat epididymal interstitium have previously been shown to move into the epididymal lumen against a concentration gradient. This is true especially in the caput epididymidis. The present investigation used the technique of in vivo epididymal perifusion and tubule micropuncture to demonstrate that the proluminal movement of 3H-androgens is subject to competitive inhibition (unlabeled testosterone in the perifusion fluid at 10 times and 100 times the concentration of 3H-testosterone significantly reduced proluminal movement of isotope) and is not energy-dependent (1 mM 2,4-dinitrophenol in perifusion fluid did not reduce the proluminal movement of isotope). Additionally, dry-mount autoradiography demonstrated high intraluminal concentrations of isotope relative to interstitial concentrations after caput tubule incubation in 3H-dihydrotestosterone (3H-DHT), and showed that the high intraluminal concentrations of isotope were not dependent on the presence of spermatozoa, i.e. proluminal movement of 3H-androgens was not due to binding to intraluminal spermatozoa. Isolation of caput epididymidal sperm on filters followed by 3H-DHT binding experiments also failed to demonstrate the presence of specific binding of this androgen to spermatozoa. Finally, it was confirmed that electrophoresed epididymal lumen fluid contains a single 3H-DHT binding peak that is at its highest concentration in the caput epididymal fluid. These data are consistent with the conclusion that intraluminal androgen-binding protein is an important factor in transepithelial androgen movement.