In this study, we compare various methods for the detection of a tumor- associated target antigen and deposition of the bound therapeutic monoclonal antibody in patients enrolled in two separate trials, one involving the administration of two radiolabeled monoclonal antibodies and the other involving an unlabeled antibody. In the first trial, patients with TAG-72 expressing metastatic colon cancer scheduled for surgical intervention received radiolabeled murine and chimeric B72.3 antibody followed by radioimmune imaging and subsequent laparotomy. Normal and tumor tissues obtained at surgery were processed for routine histology, immunohistochemistry, radiometry, and autoradiography. Both anti-TAG-72 antibodies localized to known tumor sites as evidenced by radioimmune imaging. Resected tissue revealed a high tumor-to-normal radiolocalization ratio, and autoradiography demonstrated even deposition of the radiolabeled antibodies throughout the entire tumor deposit with sparing of surrounding normal tissue. In contrast, immunohistochemistry on the same sections revealed comparatively weak antigen expression and patchy antibody localization. In the second trial, patients with GD2 antigen expressing metastatic melanoma received the unlabeled chimeric anti-GD2 antibody C14.18. Immunologic detection of the GD2 antigen and C14.18 deposition was performed on biopsy section as well as on single cell suspension. FACS analysis of the single cell suspension proved more sensitive for the detection of bound antibody than immunohistochemistry, although both methods yielded comparable results for GD2 antigen expression. Our findings demonstrate that the optimal method for the detection of tumor-associated antigen and bound therapeutic antibody can vary depending upon the nature of the antibody (radiolabeled vs. unlabeled and murine vs. chimeric), fixation stability of the target antigen, and the type of pathologic material available for study.