Overexpression of MET has been reported in various malignancies and is associated with cell proliferation, inhibition of apoptosis and metastasis. MET amplification has been reported to contribute to acquired resistance to epidermal growth factor tyrosine kinase inhibitors (EGFR-TKI) in non-small-cell lung cancer (NSCLC), being observed among 5–10% of those patients. Therefore, MET gene alterations could be both prognostic and predictive, and evaluation of that is indispensable in practice. However, the frequency of positive cases varies due to a lack of standardized criteria of fluorescence in situ hybridization (FISH). The purpose of this study is to find which criteria could be useful in terms of the association with clinical characteristics. We evaluated the MET gene copy number of FISH using two different criteria in lung adenocarcinoma: the Cappuzzo scoring system and PathVysion and its association with clinicopathological characteristics. MET positive cases according to the Cappuzzo scoring system evidenced both aneuploidy and true amplification, whereas PathVysion revealed only amplification. One hundred and thirty-eight lung adenocarcinoma tissue samples were evaluated, and the proportion of MET FISH positive cases was 15 and 4% determined by the Cappuzzo system and PathVysion, respectively. PathVysion demonstrated higher frequencies of MET FISH positives among men and smokers and evidenced no MET FISH positives in patients with bronchioloalveolar carcinoma. Prognosis was significantly associated with MET FISH positive only as defined by the PathVysion system (gene amplification), not by the Cappuzzo system. However, the progression-free survival time of patients with both EGFR mutations and MET FISH positive defined by the Cappuzzo scoring system was significantly shorter than with EGFR mutations alone. These results suggest that MET FISH is a potential prognostic factor and co-existence of MET FISH with EGFR mutations signifies worse prognosis.