The application of various types of phosphate fertilizers to sugar cane was compared in 3 experiments carried out in soil of the "massape-salmourao" type at the Usina Itaiquara, Tapiratiba. In a first trial, nitrogen and potassium were applied uniformly to all plots receiving fertilizers, and 8 different types of phosphates were compared at the rate of 120 kg of P2O5 per hectare. In two other tests a uniform rate of NK was applied, and four levels, 0, 40, 80 and 120 kg per hectare, of 5 phosphate fertilizers were compared. The results obtained in the three tests permitted the following conclusions: a) In a "massape-salmourao" soil with high acidity, a good increase in cane tonnage was obtained by the application of NK alone, when this treatment was compared with control plots receiving no fertilizer. The yield increase was considerably greater when phosphate in any of the forms was applied. There was no statistical difference in the cane yields obtained with the use of superphosphate, degelatinated and steamed bone meal, "serranafosfato", "hiperfosfato", Argelian phosphate, and a local rock phosphate (Morro do Serrote). Superphosphate and the two types of bone meal were statistically better than a type of phosphorated bauxite, but there was no difference between this and the remaining treatments. b) In "massapê-salmourão" soils of low acidity there was no statistical difference between the applications of phosphorus as superphosphate, degelatinated bone meal, "fertifós", and "hiperfosfato"; all these treatments were better than rock phosphate (Morro do Serrote).