In this paper are reported the results of 37 experiments - six repeated in the same plots for two or three consecutive years - designed to study the effect of castor pomace and cottonseed meal as fertilizers for the cotton crop. In these experiments, conducted between 1937 and 1945 and located at eleven different sites comprising the main soil types of the Planalto Paulista, were compared sources of nitrogen (castor pomace, Chilean nitrate, ammonium sulphate etc.), increasing rates of application of cottonseed meal, and methods of application of cottonseed meal. Applied by the usual method, consisting in the addition of the fertilizer to the seed furrow and mixing it with the soil just before planting, castor pomace and cottonseed meal generally gave poor results; in a few cases only were obtained satisfactory responses, but in several others they did not increase or even depressed the yield. This was observed in the absence as well as in the presence of phosphorus and potassium fertilizers. Rates of 600 and 800 kilograms of cottonseed meal per hectare generally gave no better results than 400 kilograms. Castor pomace was inferior to the soluble nitrogenous fertilizers used. The poor performance of the organic fertilizers is attributed principally to the serious injury they caused to the germinating seed and the little efficiency of the subsequent replanting of the missing hills. The method of application used in the experiments was responsible for the injury. Applied in the seed row about four weeks before planting the organic fertilizers caused no injury to the germinanting seed. Concerning the germination the placement of the fertilizers at planting time, in a furrow about four inches to one side of the seed row, was slightly inferior to the application in advance of planting. As conditions were not favorable to the utilization of the fertilizers by the plants during the season in which these two methods were studied, no definite conclusion can be drawn on their relative efficiency for the production. Nevertheless, considering some objections against the application in advance of planting and the factors that favor the side placement, it is suggested to start new experiments to study the question, including modifications to improve this method.