Lung carcinoma is the main cause of cancer death worldwide. Adenocarcinoma molecular biomarkers have been discovered, and targeted therapies have been developed with encouraging results. The epidermal growth factor receptor gene is one of these biomarkers. Exons 18 to 21 should be studied in patients with advanced adenocarcinoma, who are candidates for treatment with tyrosine kinase inhibitors. The objective was to compare the performance of the determination in large and small samples in daily practice conditions, trying to adjust to published consensus guidelines. A retrospective observational study of 141 cases was carried out, with exons 19 and 21 sequencing. Sample size (small vs. large), including number of satisfactory polymerase chain reaction (PCR), sequencing, deletions, and mutations, were evaluated. In small biopsies, sample type, fragment number, and percentage of tumor per sample were analyzed. The results shown 114/141 (80.8) cases that met selection criteria; 60/114 (53%) were large (surgical) and 54/114 (47%) were small samples (19/54 endoscopic, 17/54 fine needle aspiration clots, 4/54 lymph nodes, 14/54 core and other). All large samples were satisfactory PCR, 56/60 (93%) satisfactory sequencing, and 12/56 (21%) had deletions in exon 19. Small samples were satisfactory PCRs in 50/54 (93%) cases, and satisfactory sequencing in 35/50 (65%), 8/35 (23%) showed alterations in exon 19, and 1/35 (3%) in exon 21. In conclusion, the proportion of samples unfit for the study of the epidermal growth factor receptor gene mutational status increased from 7% in large samples to 35% in small ones. Nineteen small samples were inconclusive, with cell blocks predominating, 10/19 (53%).