Diets described in historical studies of human pellagra, have been reexamined by using modern tables of nutrient composition. Five apparently pellagragenic diets from field studies were estimated to have niacin equivalents (NE) well in excess of the Recommended Dietary Allowance (RDA), i.e., 6.6 mg/1000 kcal. In these diets, the estimated contribution of riboflavin was below the RDA, i.e., 0.6 mg/1000 kcal. A series of trials of the pellagra-preventive value of different foods carried out by the U.S. Public Health Service from 1920 to 1936 used over 400 women subjects, of whom fifty developed signs of pellagra on diets now calculated to contribute 7.2-10.6 mg NE/1000 kcal. Most of these diets apparently had less than 60% of the RDA for riboflavin. Further calculations, assuming that the niacin in ordinary cereal foods has only 30% bioavailability, still left half of the pellagragenic diets apparently with more than the RDA for "available NE." Possible reasons for the apparent discrepancy between these results and the RDA's calculated from more recent controlled studies are discussed. It is suggested that the pellagra epidemic in the U.S.A. in the early 1900's has not yet been explained satisfactorily.