Prothrombin was assayed using chromogenic substrate of S-2238 for patients who were being treated with coumarin, for patients who had liver disease, and for patients who had congenital hypoprothrombinemias and dysprothrombinemias. In coumarin therapy and in patients with liver disease the levels found correlated well with the one-stage clotting methods. The same was true for heterozygous and homozygous "true" prothrombin deficiency. In the case of congenital dysprothrombinemias the levels observed with the chromogenic substrate were higher than the clotting counterparts, particularly so in the case of prothrombin Padua. In the latter case the levels observed were always about 100% of normal, as compared with the levels of about 50% of normal found with clotting methods. These data indicate that chromogenic substrates are not always equivalent to "clotting" substrates, namely, that amidolytic activity is not always equivalent to clotting activity. Therefore the two methods cannot be used interchangeably, lest some defects escape detection.