The reliability of cyto-diagnosis and needle-biopsy in the diagnosis of malignant pleural effusions is examined on the basis of a personal series of 237 malignant pleurisies and a review of the literature. The results of the two methods, taken separately and together (respectively 42%, 46% and 56% in this study), are below the average figures obtained by compiling 18 large series in the literature (50, 60 and 70%). The reasons for this divergence in the statistics are studied, as well as the uneven reliability of cyto-histology, depending on the site of the original cancer. The authors recalled the simplicity, the reliability and, at the present time, the unrivalled usefulness of these routine methods of cyto-histology in the study of pleural carcinoma. Nevertheless, the authors also point out these routine methods and the equivocal significance of the negative result. The false security of these negative results, in the context of radical surgery for cancer, is stressed as the use in such cases of other more direct methods of exploring the pleura.