OBJECTIVE Anaplastic lymphoma kinase (ALK) gene rearrangement exists in approximately 3-7% of non-small cell lung cancer (NSCLC) and more than 15% split or isolated red signals among 50 tumor nuclei scored in the FISH analysis defines as ALK-positive. The previous studies showed that the high EGFR mutational load related to better outcomes in EGFR mutant NSCLC. Therefore, we aimed to investigate the effect of the ALK break-apart ratio on treatment outcome in metastatic ALK-positive NSCLC. METHODS The patients (pts) who ALK-positive and treated with crizotinib were retrospectively enrolled. The 30%, 40%, 50%, 60%, and 70% break-apart ratios were determined as a threshold value, and each of these was evaluated separately. Based on the results of these analyses, we detected the optimal threshold value and also performed further investigations. RESULTS A total of 70 patients were enrolled in the study. The most significant decrease in the risk of the progression or death was detected at the 50% threshold value and it was accepted as the optimal threshold. The median PFS was 17.9 vs. 7.06 months (mo) in the pts with high ALK rearrangement than low (HR: 0.43, 95% CI 0.24-0.76, p 0.004). The median OS was also significant longer in high ALK rearrange group (44.6 mo vs. 16.8 mo; HR: 0.37, 95% Cl 0.1883-0.7315; p 0.004). The intracranial progression during crizotinib treatment was significantly frequent in the pts with high ALK rearrangement (62.5-32.5%, P 0.039) DISCUSSION: In this study, we found that the high break-apart ratio can predict the extent of benefit from targeted therapy in ALK-positive NSCLC patients. Based on the results of this study, the percentage of the ALK rearrangement can be used to predict treatment outcome and to choose the optimal targeted agent in the treatment of metastatic ALK-positive NSCLC.