In the diagnosis of primary carcinoma of the lung, the history, physical examination and roentgenography are being supplemented more and more frequently by bronchoscopic examination and biopsy. This present report considers the extent to which the results of tissue examination in such cases can be of greater value to the clinician than merely to confirm the diagnosis of carcinoma. From the study of our own material and of numerous case reports it appeared that there were differences in the mode of growth and of metastasis of individual neoplasms; and that these differences had relationship to cell type. In order to investigate this, a series of 100 cases was built up, including the 34 reported from this laboratory (1). Only cases with a complete post-mortem examination and adequate microscopic control were chosen. It has been shown previously (1) that practically all bronchogenic carcinomas can be classified in three groups, depending upon the predominating cell type; the cylindric-celled or adenocarcinoma, the squamous-celled carcinoma, and the undifferentiated or small-celled carcinoma, and that these can be arranged in a scale according to their entdifferentiation without regard to histogenesis. Table I gives the sources of the 100 cases and their distribution according to cell type. The total tabulation of metastases according to cell type is shown in Table II. In presenting the tendency to metastasis, of the different cell types, the coefficient of association has been used. For a complete explanation of the derivation of the formula, the reader is referred to the original source (2). The reason for this choice rather than the use of percentages is that the numerical value Q is the result of a four-fold expression involving attribute occurrence and attribute non-occurrence.