Detailed evaluations of the microdistribution of 125I-labeled monoclonal antibodies (MoAbs) to normal tissue antigens were conducted in BALB/c mice. MoAb 273-34A, which binds to a target molecule on the lumenal surface of lung endothelial cells, localizes quickly and efficiently throughout the lung vasculature. MoAb 133-13A, which binds to an antigen on macrophage-like cells expressed in nearly equal amounts in lung, liver, and spleen, localizes most efficiently to spleen and less well to liver and lung. The microdistribution of MoAb 133-13A in liver and spleen is consistent with the antigen distribution in these organs, but in the lung a more diffuse microdistribution is observed, indicating poor access of MoAb to the antigen-positive alveolar macrophages. These findings are consistent with the hypothesis that tight endothelium (lung) represents a significant barrier to extravasation of MoAb into tissue while fenestrated (spleen) and sinusoidal (liver) endothelium are more easily penetrated. In human tumor bearing nu/nu mice, the microdistribution of MoAb to the beta 4 and alpha 6 subunits of integrin was studied. These MoAbs do not cross-react with murine integrins and thus are tumor-specific in the nu/nu mouse model. Localization of 125I-labeled MoAb 450-11A, which reacts with an intercellular domain of beta 4 integrin, is very weak and diffuse. All MoAbs to extracellular domains (mouse 450-9D, 450-30A1, and rat 439-9B) localize well to the tumor. Microdistribution of these MoAbs in the 3 different tumors is nonuniform with heavy distribution near the blood vessels, whereas antigen distribution as determined by immunoperoxidase shows a much more uniform pattern throughout the tumors. In experiments with 125I-labeled MoAb 439-9B F(ab')2, the nonuniform pattern of distribution was not changed. Gross and microdistribution of different doses of 125I-labeled MoAb 439-9B were studied. The percent of injected dose per g of MoAb in the tumor at 48 h did not vary significantly (P greater than 0.1) up to a dose of 500 micrograms/mouse, and active MoAb was recovered in comparable amounts in the serum from animals in all doses. In contrast, the microdistribution of MoAb at the high dose was different than that at low doses. At doses up to 100 micrograms/mouse, a perivascular pattern was obtained, whereas at 500 micrograms/mouse the 125I-labeled MoAb was distributed nearly evenly throughout the tumor. These data indicate that high doses of MoAb penetrate deeply into portions of the tumor that are distant from blood vessels.(ABSTRACT TRUNCATED AT 400 WORDS)