Background: There are few therapies for second-line KRASm CRC. Inhibiting downstream signal transduction may offer therapeutic options. Use of selumetinib (MEK 1/2 inhibitor; AstraZeneca) is supported by preclinical and clinical evidence. We designed a dose-finding/phase II study of IRI + SEL in KRASm CRC. Methods: Eligibility included: KRASm or BRAFm CRC with measurable disease progressing after 1st-line therapy with an oxalipatin + bevacizumab regimen; PS 0-1; acceptable organ function. Patients (Pts) were treated with IRI 180 mg/m2 iv q2w and SEL 50 or 75 mg po bid. Dose escalation was traditional 3+3 (50 mg bid SEL, then 75 mg bid). In Part B/phase II, primary endpoint was PI-determined response rate (RR) by RECIST. A Simon 2-stage design allowed expansion to 45 pts if ≥1 responses in 20 pts was seen; ≥4/45 responses would be encouraging, when compared to historical RR of 4% (and median PFS 2.5 mo) [EPIC, Sobrero 2008], with approximately 90% power to detect an ORR of 15% at the 10% alpha level (one-sided). Results: N =32 pts entered; 31 treated. Median age was 54 (27-75) yrs; 18 male and 24 Caucasian. The first 3 pts tolerated SEL 50 mg bid without DLT and the remaining 28 were treated at 75 bid. Median number of cycles on study was 3.5 and median PFS was 3.4 mo. Grade 3 AEs included (N): diarrhea 3, fatigue 2, neutropenia 2, and 1 each thrombocytopenia, enteritis, GI bleed, rash. There was one Grade 4 neutropenia. The best PI-reported response included 3 (10%) confirmed PR and 16 (52%) SD [including 1 unconfirmed PR]. 6 patients were on study for more than 6 (up to 22) months. The study was terminated early due to non-protocol considerations. Conclusions: In this small study, the RR of 10% and med PFS of 3.4 mo in pts with KRASm CRC treated with IRI + SEL in 2^nd line are promising compared with prior studies in non-selected patients. MEK inhibition in KRASm CRC should be explored further. Supported in part by AstraZeneca.