Background: Antiangiogenic therapy can rapidly reduce vascular permeability, but high doses of bevacizumab (BEV) may induce selective pressure to promote resistance. This trial evaluated the efficacy of low dose BEV in combination with lomustine (CCNU) compared to standard dose BEV in patients with recurrent glioblastoma. Methods: Eligibility criteria included age > 18 yrs with recurrent glioblastoma after temozolomide chemoradiotherapy. Patients were stratified by age, KPS, 1^st/2^nd/3^rd recurrence, and scheduled surgery and randomized to BEV 5 mg/kg every 3 weeks plus CCNU 90 mg/m² every 6 weeks or BEV 10 mg/kg every 2 weeks. The primary objective was progression free survival (PFS) with secondary objectives of radiographic response (RR), 6 month progression-free survival (PFS-6), overall survival (OS), time to progression (TTP), and safety. Results: 70 of 82 patients accrued were randomized and 68 were treated from 2010-2014. Median follow-up was 30.5 months. At the time of this analysis, 62 patients had progressed or died. For 68 evaluable patients, median PFS was not significantly longer in the BEV + CCNU arm (4.34 months, CI: 2.96-8.34) compared to the BEV alone arm (4.11 months, CI: 2.69-5.55, p = 0.19). In patients with one recurrence (n = 24 and 23, respectively), there was a trend towards statistical significance in longer median PFS time in the BEV + CCNU arm (4.96 months, CI: 4.17-13.44) compared to the BEV alone arm (3.22 months CI: 2.5-6.01, p = 0.08). Median OS in patients with 1 recurrence on BEV + CCNU was longer (13.05 months, CI: 8-19.17) than those treated with BEV alone (8.79 months, CI: 6.42-20.22, p = 0.77) but this did not reach statistical significance. No unexpected grade 3 or 4 toxicities were observed. The trial was closed early due to futility. Conclusions: The combination of low dose BEV + CCNU is a safe and effective regimen. The study was not designed to test only at first recurrence, but in that subgroup there was a strong trend towards the combination supporting the findings of the recently published BELOB trial.With a larger sample size, a statistically significant difference in median PFS between treatments arms would likely have been observed. Clinical trial information: NCT01067469