This paper reports the results obtained in two pot and eight field experiments with potatoes, which were conducted from 1943 to 1947 and designed to study the effect of increasing rates of potassium sulphate (15, 30, 45, 60, 90 and 120 kilograms of K2O per hectare) in the presence of nitrogen and phosphorus. The pot experiments were conducted in the Central Experiment Station, Campinas, and the field trials were located on different areas of four sites of the State of São Paulo. Averaging the results obtained with all the doses the responses to potash were small in both pot and field trials. In the average of the latter the maximum effect of potash, + 10%, was obtained with about 45 kilograms of K2O per hectare. In three of them, located on soils with fair to high potash contents, the addition of this nutrient did not increase the yields. In the other five experiments, some of which were located on soils low in potash, the highest response was reached also with about 45 kilograms of K2O, but the average yield increase was raised to 18%, although the stands of the experiments were seriously damaged. The injury to the stands was attributed to dry weather after planting and the addition of the fertilizers (nitrogen, phosphorus ond potash) to the furrows just before planting and slightly mixing them with the soil. It was concluded, once more, that tar evaluating the effect of fertilizers the traditional method of application must be changed.