INTRODUCTION: In randomly assigned studies with EGFR TKI only a minor proportion of patients with NSCLC have genetically profiled biopsies. Guidelines provide evidence to perform EGFR and KRAS mutation analysis in non-squamous NSCLC. We explored tumor biopsy quality offered for mutation testing, different mutations distribution, and outcome with EGFR TKI. PATIENT AND METHODS: Clinical data from 8 regional hospitals were studied for patient and tumor characteristics, treatment and overall survival. Biopsies sent to the central laboratory were evaluated for DNA quality and subsequently analyzed for mutations in exons 18-21 of EGFR and exon 2 of KRAS by bidirectional sequence analysis. RESULTS: Tumors from 442 subsequent patients were analyzed. For 74 patients (17%) tumors were unsuitable for mutation analysis. Thirty-eight patients (10.9%) had EGFR mutations with 79% known activating mutations. One hundred eight patients (30%) had functional KRAS mutations. The mutation spectrum was comparable to the Cosmic database. Following treatment in the first or second line with EGFR TKI median overall survival for patients with EGFR (n = 14), KRAS (n = 14) mutations and wild type EGFR/KRAS (n = 31) was not reached, 20 and 9 months, respectively. CONCLUSION: One out of every 6 tumor samples was inadequate for mutation analysis. Patients with EGFR activating mutations treated with EGFR-TKI have the longest survival.