Epidermal growth factor receptor (EGFR) gene mutations and increased copy numbers are considered as predictors of response to EGFR tyrosine kinase inhibitors (EGFR-TKI) in non-small-cell lung cancer (NSCLC). Lung cancer diagnosis is often based on cytology alone. However, almost all published data on EGFR gene analyses were obtained from biopsies. This study tested the feasibility of EGFR gene analyses on cytological specimens. Eighty-four cytological specimens from NSCLCs were prospectively analysed for EGFR gene mutation in exons 18-21 and EGFR gene copy numbers were evaluated by fluorescence in situ hybridisation (FISH). A FISH-positive result was defined according to the criteria by Cappuzzo et al established for biopsies of NSCLCs. Fluorescence in situ hybridisation results of cytological specimens were compared to the FISH results on matching biopsies (n=33). Initial diagnosis of NSCLC was solely based on cytology in 37 out of 84 (44.0%) patients. Out of 80 NSCLCs, 6 (7.5%) showed EGFR gene mutations. Out of 67 cancers, 45 (67.2%) were FISH positive on cytological specimens. Comparison of FISH showed a FISH-positive result in 21 out of 33 (63.6%) cytological specimens but in only 8 out of 33 (24.2%) matched biopsies. Epidermal growth factor receptor gene analyses are well applicable to cytological specimens. The high FISH-positive rate of NSCLC on cytological specimens contrasts with the low rate on biopsies when previously suggested criteria are used. New criteria for a positive EGFR FISH status to predict response to therapy with EGFR-TKI need to be defined for cytological specimens.