The present work is an attempt to correlate the results obtained in the extraction of soil phosphorus by various chemical methods with those obtained by the biological seedling method of Neubauer. The soil samples investigated represent three of the main soil types of the State of São Paulo, Brazil. Physico-chemical properties of these soils are given in the Portuguese text. Two samples of each soil type were prepared. To one of the samples superphosphate was added at the rate of 115 kg of P2O5 per hectare before being tested. Aliquots of the two samples were then extracted with the following extractors : distilled water ; 0.25 normal acetic acid ; 0.05 normal sulfuric acid ; 0.025 normai ammonium fluoride and 0.05 normal sulfuric acid ; 0.25 normal oxalic acid and 0.75 normal potassium oxalate. Rice plants (Oryza sativa L.) were used for the Neubauer test instead of rye (Secale cereale L.) because they had been found to be more satisfactory in previous tests. The results, representing the average of three replications, are summarized in table 1. Figures in column 4 represent the differences between the phosphorus absorbed from 100 g of soil and from the blank (sand). The phosphorus content of 100 rice plants that grew in sand (blank) for 17 days was 24.5 mg of P2O5. The rice plants did not take any phosphorus from the samples of the three soil types to which no superphosphate had been added. On the contrary, in two of the three soils, the plants lost some of their phosphorus, that is, phosphorus migrated from the plant to the soil colloids and became unavailable. This fact was observed for the first time in an experiment carried out in 1945/1946 in which, using the Neubauer technique with rye and rice to study 35 soil samples from the State of São Paulo, more than 70 per cent of the data obtained was negative. Figures in the other columns represent miligrams of P2O5extracted from 100 g of soil with the solutions already mentioned. The data obtained for the samples to which no superphosphate had been added indicate that the Neubauer method gave lower results than any of the chemical methods. The results for the samples mixed with superphosphate show that extraction of phosphorus from the soil was effected in the following order : distilled water <0.25 normal acetic acid < rice < 0.05 normal sulfuric acid < 0.025 normal ammonium fluoride and 0.05 normal sulfuric acid < 0.25 normal oxalic acid and 0.75 normal potassium oxalate. The Neubauer method gives results that may be regarded as low because the amount of P2O5 contained in 100 seeds of rice or rye is high (between 18 to 25 mg of P2O5). Being this amount enough for 100 seedlings during their first 17 to 18 days, they do not need to draw P2O5 from the soil. It seems, therefore, that neither rice nor rye are adequate plants to be used in the study of phosphorus intake from the soil when the Neubauer technique is employed.