FHIT is a tumor suppressor gene that is frequently silenced in non-small cell lung cancer (NSCLC) and also in preneoplastic lesions. Promoter hypermethylation was previously observed in NSCLC, and its epigenetic silencing, observed on mRNA or protein level, was proposed to predict NSCLC outcome. In the present study we evaluated the relationship between FHIT expression on mRNA level and promoter methylation, or immunoexpression level. The aim of this study was to analyze the usefulness of FHIT as early differentiating biomarker in NSCLC patients. Lung tissue specimens were obtained from 59 patients with diagnosed NSCLC (SCC=34, AC=20, LCC=5). FHIT promoter methylation was assessed in methylation-specific PCR. Relative expression analysis of FHIT was performed in real-time PCR (qPCR) and protein immunoexpression by ELISA assay. Significant differences in FHIT expression between NSCLC histopathological groups (SCC, AC, LCC) were observed (p=0.000009), with the lowest level in SCC. FHIT expression was significantly higher (p=0.034) in men vs. women. Methylated FHIT alleles were present both in NSCLC and control specimens. Mean MI value was higher in control tissue vs. neoplasm, and in men vs. women and it increased with patient age. Significant increase in MI level was observed in N0 group vs. N1 and N2, according to the TNM staging (p=0.0073). Differences in FHIT expression levels between AC, LCC and SCC indicated the usefulness of this gene as a diagnostic marker for NSCLC subtype differentiation. FHIT promoter hypermethylation both in cancer and control tissue indicated the presence of epigenetic alterations in early stage of NSCLC development. Differences in gene promoter methylation between cancer patients with and without node infiltration might be considered as a prognostic marker. Significantly lower FHIT protein immunoexpression was revealed in the group with long and intense history of smoking assessed as PYs (PY<40 vs. PY≥40, p=0.01). These results suggest the need of further study on FHIT as a potential biomarker.