The authors evaluated three methods for isolating ferritin, for use as a standard, with respect to purity of the products, ease of preparation, and yield. Examination of the respective products by gel filtration on Sephadex G 200 and Sepharose 6B suggested that the preparations isolated by ammonium sulfate and cadmium sulfate precipitation (Method 1) and by ultracentrifugation (Method 2) were homogeneous, while the product of a procedure including precipitation with ammonium sulfate (Method 3) contained significant amounts of nonferritin protein. The ratios of ferritin as measured by immunoradiometric assay to the amount of protein in the product indicated the ferritin prepared by Method 1 to be the most highly purified. Methods 1 and 2 were both comparatively simple. Although the yield from Method 1 was lowest, it is probably the method of choice, on the basis of the ease of obtaining a highly purified product. The most appropriate method for estimating protein in the isolated preparations appears to be that of Lowry et al.