Background: Gefitinib is a selective inhibitor of the epidermal growth factor (EGFR) tyrosine kinase, which is overexpressed in many cancers, including non-small-cell lung cancer (NSCLC). We carried out a clinical study to compare the relationship between EGFR gene copy number, EGFR protein expression, EGFR mutations, and Akt activation status as predictive markers for gefitinib therapy in advanced NSCLC. Methods: Tumors from 102 NSCLC patients treated daily with 250 mg of gefitinib were evaluated for EGFR status by fluorescence in situ hybridization (FISH), DNA sequencing, and immunohistochemistry and for Akt activation status (phospho-Akt [P-Akt]) by immunohistochemistry. Time to progression, overall survival, and 95% confidence intervals (CIs) were calculated and evaluated by the Kaplan-Meier method; groups were compared using the log-rank test. Risk factors associated with survival were evaluated using Cox proportional hazards regression modeling and multivariable analysis. All statistical tests were two-sided. Results: Amplification or high polysomy of the EGFR gene (seen in 33 of 102 patients) and high protein expression (seen in 58 of 98 patients) were statistically significantly associated with better response (36% versus 3%, mean difference = 34%, 95% CI = 16.6 to 50.3; P<.001), disease control rate (67% versus 26%, mean difference = 40.6%, 95% CI = 21.5 to 59.7; P<.001), time to progression (9.0 versus 2.5 months, mean difference = 6.5 months, 95% CI = 2.8 to 10.3; P<.001), and survival (18.7 versus 7.0 months, mean difference = 11.7 months, 95% CI = 2.1 to 21.4; P = .03). EGFR mutations (seen in 15 of 89 patients) were also statistically significantly related to response and time to progression, but the association with survival was not statistically significant, and 40% of the patients with mutation had progressive disease. In multivariable analysis, only high EGFR gene copy number remained statistically significantly associated with better survival (hazard ratio = 0.44, 95% CI = 0.23 to 0.82). Independent of EGFR assessment method, EGFR+/P-Akt+ patients had a statistically significantly better outcome than EGFR-, P-Akt-, or EGFR+/P-Akt- patients. Conclusions: High EGFR gene copy number identified by FISH may be an effective molecular predictor for gefitinib efficacy in advanced NSCLC.