The inaccessibility of radiolabeled antibody to poorly vascularized regions of solid tumors may reduce the therapeutic efficacy of these macromolecules. Theoretical mathematical models have predicted that increasing the protein dose administered would reduce the heterogeneity of radioantibody distribution. This investigation was undertaken to evaluate this hypothesis in experimental animal models. We have utilized the technique of macroautoradiography to demonstrate an increase in tumor penetration of the lower-affinity 125I-labeled NP-4 or higher-affinity Immu-14 anti-carcinoembryonic antigen (anti-CEA) mAbs into small (60.25-0.4 g) and large (0.8-1.5 g) GW-39 and LS174T human colonic xenografts, grown subcutaneously in the nude mouse, when 400 micrograms unlabeled antibody is administered simultaneously with 10 micrograms (100 microCi) radioantibody. Further increases in protein to 800 micrograms result in a reduction in total tumor uptake of the antibody. These in a reduction in total tumor uptake of the antibody. These differences in mAb distribution could be visualized as early as 1 day after antibody injection. Improved mAb penetration was also achieved for the Mu-9 anti-CSAp (anti-mucin) antibody using 800 micrograms unlabeled antibody. An irrelevant antibody (AFP-7-31) was found to be homogeneously distributed 3 days after injection, even at a low protein dose. Attempts to improve mAb penetration by increasing the protein dose in the GS-2 colorectal tumor, a model that has low NP-4 accretion as a result physiological barriers separating antibody from antigen, were not successful. These results suggest that a more homogeneous distribution of radioantibody can be achieved by carefully selecting a dose of unlabeled antibody to coadminister. Work is currently in progress to determine the effect of improved tumor distribution of radioantibody on the therapeutic potential of a single dose of radioantibody.