The prognostic role of epidermal growth factor receptor (EGFR) remains controversial in patients with lung cancer. Previous assays for EGFR have primarily been qualitative or, at best, semiquantitative. In the present study, using fresh-frozen tissue from 190 unselected lung cancer patients, quantification of EGFR (EGFR(ELISA)) using a recently developed enzyme-linked immunosorbent assay (ELISA) technique was compared with results (EGFR(IHC)) obtained using immunohistochemistry (IHC). Correlation between results obtained by the two different techniques was highly significant (r(s) = 0.63, P < 0.001, n = 190). This correlation improved even further (r(s) = 0.76) when sections were estimated using an IHC score that took into account percentage staining, intensity and relative tumour area. Furthermore, the relationship between clinicopathological features and prognosis was identical for the two methods. The expression of EGFR was highest in squamous cell carcinomas, but it was not correlated with other characteristics such as age, sex, histological grading, stage or prognosis. We conclude that evaluation of EGFR content using IHC and ELISA produces comparable results.