PURPOSE: Mutations of the K-ras gene were identified as a prognostic marker in metastatic colorectal cancer (mCRC). In addition, emerging data suggest that K-ras mutations are a negative predictor of clinical benefit from anti-epidermal growth factor receptor treatment in mCRC. Previously reported data suggest that the longer overall survival (OS) observed with bevacizumab treatment in mCRC is independent of alterations in the Ras/Raf/Mek/Erk pathway. We conducted additional analyses to better describe the clinical benefit of bevacizumab treatment in mCRC relative to K-ras mutation status. PATIENTS AND METHODS: Additional statistical analyses were done with data from K-ras mutation analyses in 230 patients who were treated with irinotecan, fluorouracil, and leucovorin (IFL) in combination with either bevacizumab or placebo in a randomized phase III study. Following microdissection, tissue was subject to DNA sequencing to identify K-ras mutations in codons 12 and 13. Hazard ratios for the bevacizumab group relative to the control group were estimated from an unstratified Cox regression model. The median progression-free survival (PFS), OS times, and objective response rates were compared. RESULTS: K-ras status was assessed in 230 patients (28.3%). The median PFS was significantly longer in bevacizumab-treated patients with wild-type (wt)- (13.5 versus 7.4 months; hazard ratio 0.44, p < .0001) and mutant (m)-K-ras (9.3 versus 5.5 months; hazard ratio 0.41, p = .0008). A significantly higher response rate for IFL plus bevacizumab was observed only in wt-K-ras patients (60.0% versus 37.3%, p = .006) compared with 43.2% versus 41.2% in the m-K-ras group. CONCLUSION: Bevacizumab provides significant clinical benefit in patients with mCRC expressing either mutant or wild-type K-ras.