Background: ALK rearrangement is documented in 2%-7% of NSCLC, depending on the population studied and detection method used. Although the reverse transcriptasepolymerase chain reaction (RT-PCR) was the first used and published method, fluorescence in situ hybridization (FISH) has become the primary standard diagnostic method. Recently, immunohistochemistry (IHC) has also proven to be a reproducible, faster and sensitive technique. This is one of the first studies concurrently comparing all three techniques in resected lung adenocarcinomas from the large ETOP Lungscape cohort. Methods: 95 cases from the ETOP Lungscape iBiobank, selected based on any degree of IHC staining (clone 5A4 antibody, Novocastra, UK), were examined by ALK FISH (Abbott Molecular, Inc.; Blackhall, JCO 2014) and central RT-PCR. For the latter, formalin-fixed, paraffin-embedded (FFPE) unstained slides were collected from participating centers. Slides were de-paraffinized, Toluidine Blue stained, and tumors macro-dissected. Tissue digestion and RNA extraction were performed (Qiagen RNeasy FFPE Kit). Using primers described in the literature covering most of ALK known translocations, RT-PCR (Superscript One-Step RT-PCR with Platinum Taq - 40 loops) was performed, followed by capillary electrophoresis in two separate mixes. Co-amplification of B-actin was done to validate the procedure and RNA quality. All tests were duplicated. Results: 76 of 95 RT-PCR had adequate RNA quality (B-actin co-amplification present). Among these, 18 were FISH positive, 16 were RT-PCR positive, including EML4-ALK V3a/b in 7, V1 in 5, V2 in one, and undetermined variants in 3 cases. 53 of 54 FISH negative cases were also RT-PCR negative (98%). 15 of 18 FISH positives harbored a translocation by RT-PCR (83%). Among the 4 discrepant cases, 2 FISH+/ RT-PCR- cases had IHC H-scores of 180 and 260, and 98.3% and 95% of rearranged cells by FISH, probably corresponding to variants not covered by the RT-PCR. One had an IHC H-score of 5, and 16% cells rearranged on FISH, most probably corresponding to a FISH false positive case. The last had an IHC H-score of 200, 13% rearranged cells by FISH, and, thus is defined as a false negative FISH result. Provided IHC is defined as positive by an H-score above 120, all but one case (H-Score 20, FISH and RT-PCR positive) gave concordant results by a combination of FISH and RT-PCR. Overall, using as true negative or true positive the concordant result of two of the methods, the third method is characterized by high specificity and sensitivity with corresponding values of 100/98/100% and 94/94/89% for IHC/FISH/RT-PCR, respectively. Conclusion: RTPCR is a very good tool for sorting discordant IHC/FISH cases, however, we do not recommend using this technique as single method due to the lower sensitivity of RT-PCR, as not all variants are covered, and also due to the limitations with RNA preservation.