Besults are presented of mineral fertilizer trials witht Irish potato crop in the State of São Paulo. The experiments were conducted in the flat lands of the Paraíba River banks. In this region, potato plantings are usually made from May to October, a period with little rain, :nul practically free from flood. From October to April, when good weather conditions prevail, rice is almost the only crop grown. Iu other regions of the State, potato is planted in February or in August. Chemical analyses of an average sample of the soil revealed a high organic matter content, about seven fold higher total nitrogen content and a twelve fold higher total carbon content, as compared with those found in a soil sample taken from Capão Bonito, an important potato area in the State, where phosphoric acid has been found to be responsible for increases in tuber productions. The experimental design was arranged so that conclusions could be drawn from the influence of increasing amounts of each nutrient element N, P and K, in the presence of a constant amount of the other two. In spite of the high total nitrogen content of the soil, results have shown this element not to he in available form for plants. In these experiments, application of nitrogen has been found to be effective and even decisive in increasing tuber yield. Lack of this element results in a lower production, in some cases, even inferior to check plots. In experiment n.° 38, for instance, PK yielded two tons of tubers per hectar, whereas NK and NP yielded, respectively, 1.1.9 and 12.4 tons. In another trial, n.° 40 BK yielded 10.5 tons per hectar (less than check), while NPK yielded 16.2 tons. From the economical point of view, an amount of 80 kg of N to the hectar in the presence of PK, brougth better profits, application of PK alone, resulted in losses. Determinations of starch content in tubers of every treatment have shown no appreciable differences between them.