Background: In the setting of metastatic colorectal cancer (CRC), anti-EGFR antibodies are not currently recommended for individuals with KRAS mutant tumours. This is based on subgroup analyses of individual clinical trials rather than a formal synthesis of evidence for KRAS status as a predictive biomarker, while newer trials report no benefit for anti-EGFR antibodies irrespective of KRAS status. This study systematically reviewed the evidence for KRAS mutation status as a treatment effect modifier of response to anti-EGFR antibodies and the influence of partner chemotherapy. Methods: Medline (1966-2010), EMBASE and American and European oncology meeting abstracts were searched for randomised controlled trials reporting the influence of KRAS status on effectiveness of anti-EGFR antibodies in metastatic CRC. The treatment efficacy was summarised by KRAS status using hazard ratios (HR) for progression-free survival (PFS) and risk differences (RD) for objective response. For each study, a measure of effect modification was calculated, and aggregated using random effects meta-analysis to assess the interaction between KRAS and treatment effect. Findings: Eleven studies (8924 patients) were sele cted from 198 reports. Two studies assessed anti-EGFR antibodies as monotherapy and nine their use with chemotherapy. KRAS status was reported in 7555 cases. In subgroup analysis, the progression HR for KRAS wild patients assigned to anti-EGFR antibodies was 0.80 (4436 patients 95%CI: 0.64, 0.99) and for mutant cases 1.11 (3119 patients, 95%CI: 0.97, 1.27). A significant treatment effect interaction between KRAS status and addition of anti-EGFR antibodies to standard treatment was found for PFS (ratio of HRs 0.71, 95%CI: 0.57, 0.90 p = 0.005) and response rate difference (difference in RDs 15%, 95%CI: 8, 22%, p < 0.001). There was no evidence that the extent of effect modification differed between chemotherapeutic partners for both PFS (p = 0.3) and response rate (p = 0.6). Interpretation: KRAS mutation status is a treatment effect modifier for anti-EGFR antibodies in metastatic CRC. Further evidence is needed to determine whether this is true for all chemotherapy partners and all clinical circumstances.