An experiment to determine the effect of four levels of N, P, K, and micronutrients on sweetpotatoes was conducted in an Oxisol in northwestern Puerto Rico. Treatment differentials included banded, postplant N at the rates of 0, 39, 78, and 157 kg/ha; Pat the rates of 0, 17, 33, and 67 kg/ha; K at the rates of 0, 67, 134, and 269 kg/ha; and a micronutrient mixture at the rates of 0, 33, 67, and 134 kg/ha. The soil was high in organic matter and fairly well supplied with N, Ca, K, and Mg. It was apparently low in extractable P. In general, the amounts of a given nutrient absorbed by the plants were not consistently related to the amount of nutrient applied, except in the case of N. Apparent fertilizer recovery was very low in all cases. Nonfertilized plots produced only 9.7 tons of marketable roots, while yields of 16 t/ha were obtained with the application of 39 kg/ha of N. The Capo fertilizer-yield equation appears to be useful in predicting sweetpotato yields in terms of the N fertilizer applied in this Oxisol. High N levels tend to induce vine growth at the expense of edible roots. Marketable root yields were not affected by P, K, and micronutrient applications.