The aim of this phase II trial was to estimate the objective response rate (ORR) of two different schedules of ixabepilone [weekly or every 3 weeks (Q3W)] combined with bevacizumab, relative to a reference arm of weekly paclitaxel and bevacizumab. Patients with human epidermal growth factor receptor 2-normal, chemotherapy-naïve metastatic breast cancer (MBC) were randomized 3:3:2 to ixabepilone 16 mg/m(2) weekly plus bevacizumab 10 mg/kg Q2W (Arm A: n = 46); ixabepilone 40 mg/m(2) Q3W (reduced to 32 mg/m(2) after four cycles of treatment) plus bevacizumab 15 mg/kg Q3W (Arm B: n = 45); or paclitaxel 90 mg/m(2) weekly plus bevacizumab 10 mg/kg intravenous infusion Q2W (Arm C: n = 32). Of 123 randomized patients, 122 were treated. All were followed for ≥19 months; 5 % of patients remained on study treatment at the time of this analysis. Grade 3 or 4 neutropenia was more common in Arm B (60 %) than Arms A (16 %) or C (22 %); other adverse events were similar. The investigator-assessed ORR was 48, 71, and 63 % for Arms A, B, and C, respectively. Median progression-free survival (randomized patients) was 9.6 months in Arm A, 11.9 months in Arm B, and 13.5 months in Arm C. In conclusion, ixabepilone Q3W plus bevacizumab has clinical activity as first-line therapy for MBC relative to paclitaxel plus bevacizumab, but with significantly greater risk of grade 3 or 4 neutropenia. In addition, these data suggest that weekly dosing of ixabepilone may be less active than Q3W dosing, but with less neutropenia.