In an effort to identify the site of recurrent colorectal cancer in patients with occult metastasis and increasing serum CEA levels, we conducted two trials using monoclonal antibodies (MoAb) against CEA. The first utilized Indium-111-labeled ZCE-025; an immunoglobulin G1 (IgG1) anticarcinoembryonic antigen (anti-CEA) antibody (Hybritech, San Diego, CA). The second study used Tc-99m-labeled Fab' fragment of IMMU-4 (Immunomedics, Morris Plains, NJ). Eighteen patients were imaged with the ZCE-025 and 14 with the Tc-99m Fab' IMMU-4. True-positive scans, defined as at least one correct correlation of the MoAb scan and surgical/histologic findings, were observed in 12 of 15 patients undergoing exploration or biopsy using the ZCE-025 and 11 of 14 using the IMMU-4. There were no true-negative scans with the ZCE-025 and only 2 of 14 with the IMMU-4. There were 3 false-positive scans with the ZCE-025 and 1 of 14 with IMMU-4. There were no false-negative scans with either ZCE-025 or IMMU-4. Four (31%) of 13 patients undergoing exploration and imaged with ZCE-025 and 5 (36%) of 14 imaged with IMMU-4 had complete tumor resection. Treatment decisions were affected in 3 (16%) of 18 ZCE-025-imaged patients and 3 (21%) of 14 IMMU-4 ones. Two (14%) of 14 patients imaged with IMMU-4 had negative MoAb scans and negative laparotomies. Despite these findings, it is not known whether such early detection and resection will translate into improved survival rates. The authors suggest two randomized studies, one designed to ascertain the role of MoAb added to blind exploratory laparotomy. In that study, patients with increasing CEA levels and a negative workup will be randomized to an exploratory laparotomy preceded by MoAb anti-CEA scans or a straight exploratory laparotomy without the assistance of a MoAb anti-CEA scan. Endpoints will be differences in complete resectability and survival. A second study will examine the merits of blind exploratory laparotomies. In that study, patients with increasing CEA levels and a negative workup would be randomized to MoAb imaging, exploratory laparotomy, and radioimmunoguided surgery, and the other cohort of patients would continue to have conventional radiologic workup. Exploration in this latter group would be performed only when indicated by radiologic or endoscopic studies. The endpoint of the study would compare survival in the two cohorts of patients. These two studies may ultimately settle the debate regarding the correct approach to patients with occult metastatic colorectal cancer and a increasing levels of serum CEA.