In view of the need for higher rates of potash fertilizers for sugar cane, a preliminary experiment was conducted to study some methods of application. Rotes of applications of 90, 180 and 270 kilograms of K2O to the hectare were compared in the presence of NP, each rote being applied in three ways: according to the usual method, i. e. the whole quantity in the furrows, at planting time; half the quantily at planting and half two months later; one third ot planting and each of the other two thirds two ond 8 1/2 months afterwards. In São Paulo the cane cuttings are planted in deep furrows ond slightly covered with soil, the filling of the furrows being achieved gradually, with the usual cultivations. The first post-planting application was made in those semi-filled furrows; the second, 20 centimeters to the side of the rows, as the furrows were almost levelled. The average response to potash was very high, but the methods of application did not differ significantly. Nevertheless, it was observed that the higher rotes were more effective when sptitted, and the advantoge of splitting increased as the total dose was increased. It seems that no appreciable losses by leaching of the doses applied at planting ocurred. On the other hand, the plants could utilize the post-planting fractions, for they were present in soil layers where roots developped before the period of intensive uptake of nutrients. Besides reducing possible losses by leaching in certain soils, sppliting of higher rates of potash would contribute to ovoid excessive salt concentration in the planting furrows. Although some symptoms of salt-injury were observed, the damage was not serious in the present experiment, because it rained before and soon after planting.