Background: KRAS mutations are among the most common genetic alterations in NSCLC; however no targeted therapies have been approved to benefit this lung cancer subset. Between 2/2008 and 2/2011 our center conducted two consecutive multicenter randomized phase II trials in patients (pts) with refractory NSCLC comparing erlotinib/placebo versus erlotinib + either sorafenib or pazopanib, both oral multikinase inhibitors (Spigel et al, JCO 2011; Chicago MSTO 2012). Progression-free survival (PFS) was improved with the multikinase regimens in the EGFR wild-type (WT) subsets, but not in the overall populations. An unplanned analysis of the combined KRAS subset data is the subject of this report. Methods: Eligibility criteria for both trials included: stage IIIB/IV NSCLC; 1 to 2 prior regimens; ECOG performance status 0–2; measurable disease. PFS was the primary endpoint of each trial. Treatment groups included: erlotinib/placebo (N=121), erlotinib/sorafenib (N=112), and erlotinib/pazopanib (N=127). 168 pts (47%) in these three groups had sufficient tumor specimens for KRAS analysis. Results: The PFS and OS results based on KRAS results are shown in the Table below. Conclusions: Patients in whom the KRAS mutation status was known achieved a significantly longer PFS with erlotinib and a multikinase inhibitor than with erlotinib alone. Although this unplanned combined analysis has several limitations, the greater PFS and OS benefits in pts with KRAS mutations warrant further study. Clinical trial information: NCT00600015; NCT01027598.