In spite of several extensive ultracentrifugal studies of human sera by Mutaenbecher, McFarlane, and Pedersen, the interpretation of the patterns observed has remained in doubt. Specifically, the greatest difficulties encountered by previous workers have arisen in their efforts to study undiluted human sera. McFarlane(1) referred to marked distortions in the pattern observed with undiluted human serum and suggested a trial and error dilution of the serum with salt solution to minimize such distortions. Pedersen (2) recommended diluting sera with various salt or buffer solutions to 40 per cent of the initial concentration, since under these conditions “adequate” resolution of the albumin and so called “X protein” peaks in the sedimentation diagram could be made. Both these workers found the apparent concentration of the X protein to vary considerably with respect to both over-all protein and salt concentration. The variations in concentration ranged from a value of X protein constituting up to 30 per cent of the serum proteins when studied in concentrated serum to an immeasurably small value when the serum was greatly diluted. Pedersen has explained this variation in X protein concentration by assuming this molecule to be a labile complex of albumin, globulin, and lipides which dissociates with increasing dilution of the serum. On the basis of ultracentrifugal studies of human sera by the present authors, a wholly different interpretation of the ultracentrifugal patterns observed is given herewith. This interpretation indicates that the X protein concentration in human sera is vastly smaller than reported by Pedersen or McFarlane, but more consistent with electrophoretic data for the low density B, lipoproteins. Further, the ultracentrifugal pattern observed for human serum with increasing dilution can be explained without assuming that any dissociation of X protein occurs.