Although most studies have indicated that Ber-EP4 immunostaining can assist in differentiating epithelial pleural mesotheliomas from adenocarcinomas that metastasize to the pleura, the percentage of positive cases has varied greatly among different studies. Authors of a recent publication concluded that Ber-EP4 has no diagnostic utility in separating these conditions. To determine whether Ber-EP4 has any value in distinguishing mesothelioma from adenocarcinoma, 70 formalin-fixed epithelial pleural mesotheliomas, 20 pulmonary adenocarcinomas, 59 nonpulmonary adenocarcinomas, 4 squamous cell carcinomas of the lung, 6 transitional cell carcinomas, and 31 adenocarcinomas of unknown origin that metastasized to the pleura were stained with this antibody. Reactivity was observed in 18 (26%) of 70 mesotheliomas and in all 20 (100%) of the pulmonary adenocarcinomas, in 55 (93%) of the 59 nonpulmonary adenocarcinomas, in 4 (100%) of 4 squamous cell carcinomas of the lung, in 4 (67%) of 6 transitional cell carcinomas, and in 26 (84%) of 31 adenocarcinomas of unknown origin that metastasized to the pleura. The staining in the mesotheliomas was focal and restricted to a limited number of cells, in contrast with staining in the pulmonary adenocarcinomas in which it was invariably diffuse. The extent of the staining in the nonpulmonary adenocarcinomas and the metastatic adenocarcinomas of unknown origin was less consistent--negative or focal in some cases and diffuse in others. Therefore, while Ber-EP4 seems to be helpful in separating epithelial pleural mesotheliomas from lung adenocarcinomas, its value in distinguishing mesotheliomas from other tumors metastatic to the pleura is more limited and depends largely on the site of origin of the metastatic tumor.