A syngeneic anti-idiotype monoclonal antibody (MAb) (CM-I I) directed against an anti-carcinoembryonic antigen (CEA) murine MAb (NP-4) was evaluated as a second antibody (SA) to promote the rapid clearance of radiolabeled NP-4 from the blood. Initial studies confirmed that CM-I I IgG removed l3Il-NP-4 IgG from the blood as effectively as a polyclonal donkey anti-goat IgG removed I3'1-goat IgG. However, use of an F(ab'), in place of either the NP-4 or CM-I I IgG was not as effective in removing primary radiolabeled antibody, despite the formation of high-molecular-weight complexes. In accordance with previous results, the timing and dose of the SA injection was critical for optimizing tumor uptake and improving tumor/non-tumor ratios. In nude mice bearing GW-39 human colonic tumor xenografts, a delay in the injection of CM-I I by 48 hr after injection of radiolabeled NP-4 was optimal, since this allowed maximum tumor accretion. At a 2001 CM-I I:NP-4 ratio, tumor uptake was reduced, suggesting inhibition of NP-4 binding to CEA within the tumor. Despite optimizing tumor uptake by delaying SA injection and adjusting i t s dose, the percentage of l3'I-NP-4 in the tumor decreased 2-to 3-fold within 2 days after CM-I I injection. A similar effect was seen for "'In-labeled NP-4 IgG with CM-I I. Injection of excess unlabeled NP-4 given to block CM-I I shortly after i t s injection failed to curtail the loss of NP-4 from the tumor. Our results suggest that high blood levels of MAb are important for sustaining NP-4 in the tumor. Radiation-dose predictions derived from biodistribution studies indicate that a higher tumor dose may be delivered using the SA method than with either I3'1-NP-4 IgG or F(ab'), alone. Use of the SA method with 9oY-labeled NP-4 IgG, as modeled from biodistribution studies with "'In-NP-4 IgG, would likely be limited by liver toxicity.