Improvements in efficacy of radioimmunotherapy will require increased tumor uptake relative to normal tissue. We previously demonstrated that labeling the IgG2b glioma-reactive antitenascin monoclonal antibody 81C6 with 131I using N-succinimidyl-3-(tri-n-butylstannyl)benzoate (ATE) increased tumor uptake and tumor-to-normal tissue ratios and decreased deiodination compared with labeling using Iodo-Gen. The present study was conducted to determine whether 131I 81C6 labeled using ATE (81C6 ATE) would demonstrate a therapeutic advantage over 131I 81C6 labeled using Iodo-Gen (81C6 IOD) in treating s.c. D-54 MG human glioma xenografts in athymic mice. The subclass IgG2b MAb 45.6 labeled with 131I using ATE (45.6 ATE) was used as a control. Animals were injected with saline or 500 microCi of 45.6 ATE (23 microCi/microgram), 81C6 ATE (26 microCi/microgram), or 81C6 IOD (24 microCi/microgram). With approximately 150 mm3 initial tumor volumes, growth delay for 81C6 ATE was significantly better by Wilcoxon rank sum analysis than saline (P = 0.0006 to 0.003), 45.6 ATE (P = 0.0006 to 0.002), and 81C6 IOD (P = 0.0008 to 0.007). Biodistribution data from similarly injected animals gave estimated radiation doses to tumor of 7723, 5200, and 1667 rad for 81C6 ATE, 81C6 IOD, and 45.6 ATE, respectively. In addition, 81C6 ATE administered at this dosage to animals with 50% larger initial tumors also improved tumor growth delay in comparison with 81C6 IOD given to animals with standard-size tumors. A similar experiment was conducted at 1000 microCi and, although radiation toxicity was noted in all labeled groups, two animals in the 81C6 ATE group had tumor regression for more than 240 days, and the other groups had no regressors. We conclude that the use of the ATE method may significantly improve the therapeutic efficacy of radioiodinated monoclonal antibodies.