Previously, this laboratory reported that in mice pre-targeted with unlabelled streptavidin, the biodistribution of 111In administered on one biotin derivative (EB1) was superior to that of another derivative (DB2). In addition, a Scatchard analysis showed that the affinity constant of 111In-EB1 is lower by seven orders of magnitude from that of 111In-DB2. Therefore, this paper considers the role that endogenous biotin may play in these observations. Both 111In-labelled EB1 and DB2 were bound to streptavidin and incubated at 37 degrees C in mouse blood with increasing concentrations of d-biotin. As determined by Sephadex G-50 chromatography, only an 8-fold molar excess of d-biotin relative to labelled streptavidin was required to displace 90% of label in the case of EB1, whereas even a 20-fold molar excess provided no detectable displacement of DB2. That this displacement was also occurring in vivo was established in a mouse model bearing an infected thigh: increasing the serum biotin level (by intraperitoneal administration of d-biotin) had no effect on the biodistribution of 111In when administered on DB2; however, the target to non-target ratio decreased in the case of EB1. We have also observed that the biodistribution is no longer favourable when EB1 is administered radiolabelled with 99Tcm. When 111In was substituted with 99Tcm on EB1, chromatography of blood samples showed that similar displacement was occurring; however, in this case, the displaced label bound to serum proteins.(ABSTRACT TRUNCATED AT 250 WORDS)