Human erythrocytes transformed leukotriene A4 into leukotriene B4. Metabolism was proportional to the erythrocyte concentration, even at subphysiological levels (0.08-4 X 10(9) erythrocytes/ml). Comparative metabolic studies excluded the possibility that leukotriene B4 originated from trace amounts of polymorphonuclear leukocytes or platelets present in the purified erythrocyte suspensions. For example, suspensions of isolated platelets (100-500 X 10(6) cells/ml) failed to convert leukotriene A4 into leukotriene B4; and conversion by suspensions of isolated polymorphonuclear neutrophils was insufficient to account for the amounts of leukotriene B4 formed by erythrocytes. Leukotriene B4 formation was maximal within 2 min and substrate concentration dependent. Enzymatic activity originated from a 56 degrees C labile nondialyzable (Mr greater than 30,000) soluble component in the 100,000 X g supernatant obtained from lysed erythrocytes. In contrast to the contemporary view, our results indicate that human erythrocytes are not metabolically inert in terms of eicosanoid biosynthesis. The role of human erythrocytes during inflammatory or pulmonary disorders deserves re-examination in this context.