The recent discovery and characterization of an oncogenic ROS1 gene fusion in a subset of lung cancers has raised significant clinical interest because small molecule inhibitors may be effective to these tumors. As lung cancers with ROS1 rearrangements comprise only 1-3% of lung adenocarcinomas, patients with such tumors must be identified to gain optimal benefit from molecular therapy. Recently, immunohistochemical analyses using a novel anti-ROS1 rabbit monoclonal antibody (D4D6) have shown promise for accurate identification of ROS1-rearranged cancers. To validate this finding, we compared the immunostaining results of tissue microarrays (TMAs) containing 17 ROS1-rearranged and 253 ROS1-non-rearranged lung carcinomas. All 17 ROS1-rearranged cancers showed ROS1 immunoreactivity mostly in a diffuse and moderate-to-strong manner with an H-score range of 5-300 (median, 260). In contrast, 69% of ROS1-non-rearranged cancers lacked detectable immunoreactivity, whereas the remaining 31% showed reactivity mainly in a weak or focal manner. The H-score for the entire ROS1-non-rearranged group ranged from 0 to 240 (median, 0). The difference in H-score between the two cohorts was statistically significant, and the H-score cutoff (≥150) allowed optimal discrimination (94% sensitivity and 98% specificity). Similar but slightly less-specific performance was achieved using the extent of diffuse (≥75%) staining or ≥2+ staining intensity as cutoffs. CD74-ROS1 and EZR-ROS1 fusions were significantly associated with at least focal globular immunoreactivity and plasma membranous accentuation, respectively, and these patterns were specific to ROS1-rearranged cases. Although full-length ROS1 is expressed in some ROS1-non-rearranged cases, we showed that establishment of an optimal set of interpretative criteria makes ROS1 immunohistochemistry a valuable method to rapidly and accurately screen lung cancer patients for appropriate molecular therapy.