Histological diagnosis of malignant mesothelioma and differentiation from adenocarcinoma is often difficult. Definitive pathological confirmation of malignant mesothelioma requires demonstration of an appropriate immunohistochemical phenotype. Selection of an optimum panel of immunohistochemical antibodies for the reliable identification of malignant mesothelioma is hindered by the absence of a specific immunohistochemical label for mesothelioma cells. Recently, we have found that the ovarian carcinoma cell antibody CA125 labels malignant mesothelioma cells, and the antibody HBME-1 has been developed as a sensitive mesothelial cell marker. We have compared the immunohistochemical staining patterns achieved with CA125 and HBME-1 to those obtained using a panel of eight further antibodies in 17 malignant mesotheliomas and 14 primary and secondary adenocarcinomas within lung and pleura. CA125 labelled malignant mesothelioma cells in 15 of 17 cases (88%), and adenocarcinoma cells in seven of 14 cases (50%). HBME-1 labelled mesothelioma cells in all 17 cases (100%) but also labelled adenocarcinoma cells in 10 of 14 cases (71%). BerEP4 positively labelled one malignant mesothelioma but was negative in the remaining 16 cases and positively labelled nine of 14 adenocarcinomas (64%). Monoclonal anti-CEA, AUA-1, CA19.9 and LeuM1 labelled no malignant mesotheliomas and were positive in 10 (71%), nine (64%), eight (57%) and six (43%) of 14 cases of adenocarcinoma, respectively. Diastase-PAS staining detected neutral mucin in none of the malignant mesotheliomas but in 10 (71%) of the 14 adenocarcinomas. We conclude that CA125 and HBME-1 do not label mesothelial cells with sufficient specificity to be useful for differentiating malignant mesothelioma from adenocarcinoma, although negative staining with HBME-1 makes a diagnosis of malignant mesothelioma unlikely. As there remains an absence of a specific positive mesothelial cell marker this distinction is still most reliably made using a panel of antibodies including at least two of the following: anti-CEA, AUA-1, BerEP4, LeuM1 and CA19.9, in combination with histochemical assessment of neutral mucin production.