Background: Patients (pts) with unresectable WT KRAS mCRC benefit from fluoropyrimidines (FP), oxaliplatin (O), irinotecan (I), bevacizumab (Bev) and E. The most cost-effective strategy to combine them remains unclear. Methods: A Markov model was constructed for a hypothetical cohort of pts with mCRC to examine the costs and outcomes of 3 treatment strategies: (A): 1st line (1L) Bev+FP+O/I, 2nd line (2L) FP+I/O, 3rd line (3L) E, (B): 1L Bev+FP+O/I, 2L FP+I/O, 3L E+I, and (C): 1L E+FP+O/I, 2L Bev+FP+I/O, 3L best supportive care (BSC). Efficacy and probability data of the treatments were obtained from clinical trials identified through a systematic review of the literature. Resource utilization data were derived from a chart review of 65 consecutive pts treated at Odette Cancer Centre (OCC) since 2009 and from the literature. Utilities were obtained by surveying oncologists (n= 24) across Canada using EQ-5D. Costs were obtained from the Ontario Ministry of Health and Long/Term Care, Ontario Case Costing Initiative, OCC and the literature. The analysis was conducted from the Canadian public healthcare system perspective over a 5 year time horizon with a 5% discount in 2012 Canadian dollars (CAD$) for cost and outcome. Incremental cost-effectiveness analyses were conducted comparing costs and outcomes of the 3 strategies. One way and probabilistic sensitivity analyses (SA) were conducted (n=10,000). Results: All 3 strategies appeared to be of relatively similar efficacy clinically, but C is more expensive than A or B by >45% (see Table). The model is primarily driven by the acquisition cost of drugs. B is most cost-effective when the willingness-to-pay (WTP) threshold >$120,000/QALY. SA showed that C would be cost-effective only when the progression-free survival of E is better than Bev in 1L with hazard ratio <0.24 at WTP of $150,000/QALY. Conclusions: 1L use of E followed by 2L Bev in mCRC is not cost-effective at the current pricing of E relative to Bev.