Background Bevacizumab significantly improves progression-free survival (PFS) and overall survival (OS) when added to chemotherapy for metastatic colorectal cancer (mCRC). The hypothesis that bevacizumab discontinuation could lead to an angiogenesis flare and eventually to an accelerated tumor progression has not been confirmed in a recent large pooled analysis. Therefore the optimal duration of bevacizumab still remains undefined. Patients and methods PFS and OS were retrospectively analyzed in patients with previously untreated mCRC who received bevacizumab 5 mg/Kg and standard FOLFIRI regimen (leucovorin, infusional fluorouracil and irinotecan) up to a maximum of 12 cycles. Results Data from 209 patients were collected and analyzed. The median follow-up was 24 months. Fifty-five (26.3%) patients received at least 6 administrations and 114 (54.5%) received a maximum of 12 administrations of bevacizumab. Median exposure to bevacizumab was 148 days (4.9 months). Median PFS and OS were 10.7 months [95% confidence interval (CI) 9.2-12.2 months] and 31.6 months (95% CI 25.8-37.3 months), respectively. Overall objective response rate was 49.8% (95% CI 42.9-56.6) and the disease control rate 81.8%. Approximately 65% and 30% of patients received some form of second- and third-line therapy, respectively. The toxicity profile of bevacizumab was consistent with that documented in previous trials. Conclusions In this retrospective analysis remarkably long PFS and OS were obtained with a first-line therapy duration limited to a maximum of 12 cycles. Our data does not support a decreased PFS or increased mortality after discontinuation of bevacizumab in mCRC patients.