Due to limited information on sweet potato (Ipomoea batatas L.) cultivation in uncultivated savanna areas, the objective of this research was to determine the ideal dose and type of organic fertilizer for sweet potato cultivation in savanna soils with no history of use. In four experiments, the following fertilizer doses were tested: cattle manure (0, 10, 20, 30, and 40 t∙ha-1), poultry manure (0, 5, 10, 15, and 20 t∙ha-1), compost A (0, 0.75, 1.5, 2.25, and 3.0 t∙ha-1), and compost B (0, 0.75, 1.5, 2.25, and 3.0 t∙ha-1). The fifth experiment consisted of an organic fertilizer efficiency test using applications already recommended for sweet potato crops, which resulted in six treatments: control (without organic fertilization), cattle manure (20 t∙ha-1), poultry manure (10 t∙ha-1), compost A (1.5 t∙ha-1), compost B (1.5 t∙ha-1), and compost C (0.375 t∙ha-1). In the first four experiments, the products and their recommended doses were defined in order of production efficiency, as follows: poultry manure (doses between 13 and 20 t∙ha-1) > cattle manure (doses between 30 and 40 t∙ha-1) > compost B (doses between 0.75 and 2.25 t∙ha-1) = compost A (doses between 2.25 and 3.00 t∙ha-1). The fifth experiment concluded that: 1) poultry manure was the most suitable starting point for sweet potato cultivation in savanna soils and 2) sweet potato yield was directly linked to the commercial root mass, number of commercial roots, and branch productivity, which, in turn, were maximized by an increase in organic matter and satisfactory amounts of phosphorus in the soil.