(1) Background: The C-ros oncogene 1 (ROS1) gene translocation is an important biomarker for selecting patients for crizotinib-targeted therapy. The aim of this study was to understand the incidence, diagnostic algorithm, clinical course and objective response to crizotinib in ROS1 translocated lung non-small cell lung cancers (NSCLCs) in Taiwan. (2) Methods: First, we retrospectively studied the ROS1 status in 100 NSCLC samples using break-apart fluorescent in situ hybridization (FISH) and immunohistochemical (IHC) staining to establish a diagnostic algorithm. Then, we performed routine ROS1 IHC tests in 479 NSCLCs, as crizotinib was available from 2018 in Taiwan. We analyzed the objective response rate and the survival impact of crizotinib. (3) Results: Four ROS1 translocations were clustered in epidermal growth factor receptor (EGFR) wild-type adenocarcinomas but not in cases with EGFR mutations. Strong ROS1 expression was positively correlated with ROS1 translocation (p < 0.001). NSCLCs with ROS1 translocation had a poor prognosis compared to those without ROS1 translocation (p = 0.004) in the pre-crizotinib stage. Twenty NSCLCs were detected with ROS1 translocation in 479 wild-type EGFR specimens from 2018. Therefore, the incidence of ROS1 translocation is approximately 4.18% in EGFR wild-type NSCLCs. In these 20 ROS1 translocation cases, 19 patients received crizotinib treatment, with an objective response rate (ORR) of 78.95% (confidence interval = 69.34% to 88.56%), including 1 complete response, 14 partial responses, 3 stable cases and 1 progressive case. Overall survival and progression-free survival were better in the 19 ROS1-translocated NSCLCs of the prospective group with crizotinib treatment than the four ROS1-translocated NSCLCs of the retrospective group without crizotinib treatment. (4) Conclusions: ROS1-translocated NSCLCs had a poor prognosis and could have a beneficial outcome with crizotinib.