PURPOSE: To clarify the effect of bevacizumab on NK012 therapy in mice bearing U87MG glioblastoma orthotopic xenografts in comparison with the combination therapy of irinotecan hydrochloride (CPT-11) with bevacizumab. EXPERIMENTAL DESIGN: NK012 at 7-ethyl-10-hydroxycamptothecin (SN-38) equivalent dose of 30 mg/kg was administered intravenously three times every 4 days with or without bevacizumab. CPT-11 at 66.7 mg/kg was administered intravenously three times every 4 days or CPT-11 at 40 mg/kg/d over 5 consecutive days with or without bevacizumab. Bevacizumab was administered intraperitoneally six times every 4 days in each experiment. In vivo antitumor effects were evaluated by bioluminescence imaging, histopathologic evaluation, and immunohistochemistry. To evaluate interaction with bevacizumab, free SN-38 concentration in tumor tissues was examined by high-performance liquid chromatography. RESULTS: CPT-11 in combination with bevacizumab showed significantly more potent antitumor activity and longer survival than CPT-11 monotherapy (P < 0.05). However, there was no difference between NK012 monotherapy and NK012 in combination with bevacizumab. Concentration of free SN-38 released from NK012 in tumor tissue decreased in combination with bevacizumab (P = 0.027). NK012 monotherapy or NK012 with bevacizumab showed potent antitumor activity and longer survival than any dosing method of CPT-11 in combination with bevacizumab (P < 0.05). Orthotopic tumors treated with NK012 showed decreased tumor cellularity and lower Ki-67 index (P < 0.001) relative to those treated with CPT-11/bevacizumab. CONCLUSIONS: The present study using orthotopic glioblastoma model in mice may warrant further preclinical evaluation of NK012 before conducting the clinical trial of the drug, because the antitumor activity of NK012 monotherapy was superior to the combination therapy of CPT-11 with bevacizumab.