This paper reports the results obtained in gix trials conducted in different zones of the State of São Paulo and designed to study mainly the effect of under-sowing the com crop, grown without rotation, with velvet beans (Stizolobium sp.). While one of the trials, at Piadamonhangaba, was interrupted after the first crop, the others, located at Capão Bonito, Pindorama. Tatuí, Limeira, and Engenheiro Hermilo, were conducted respectively for 10, 10, 9, 5 and 3 years. The treatments compared were the following: none; ground limestone; phosphorus and potassium; limestone, phosphorus and potassium; velvet beans; limestone and velvet beans; phosphorus, potassium and velvet beans; limestone, phosphoruB, potassium and velvet beans. The phosphorus and potassium fertilizers were applied annually in the furrows opened for the corn seed; the limestone was applied also annually and in the mentioned furrows, except in the last three years of the trials of Capão Bonito, Pindorama and Tatuí, when it was broadcast. The velvet beans were sown in a row between each pair of corn rows, after the beginning of the taaseling stage, and were turned under with the stovers few weeks after the harvest of the corn ears. In most of the experiments the corn yields were fairly to highly increased by the application of phosphorus plus potassium. The response to limestone was generally very small when judged by the averages of the whole period of each trial; it was observed, however, that in the longer trials it tended to grow better as the years passed, mainly in the presence of velvet beans. Excluding Pindamonhangaba, whose trial was conducted for only one season, the under-sowing of velvet beans increased significantly the corn yields at Tatuí, Capão Bonito and Pindorama, but not at Limeira and Engenheiro Hermilo. In the general averages of each of the experiments of Tatuí, Capão Bonito and Pindorama the response to the addition of velvet beans was practically the same in the absence or presence of either limestone or phosphorus plus potassium; during the last years, however, it was fairly better in the presence of both the amendment or the fertilizers. The application of these near the corn but far from the velvet bean plants seems to have delayed considerably their favorable influence on the effect of the leguminous. At Engenheiro Hermilo the velvet beans failed to grow, apparently because the soil was poor, the leguminous was not directly fertilized and the experiment was of short duration. The velvet beans grew fairly well at Limeira; but this trials was located on a pasture and it seems that the soil was well supplied with nitrogen. The results obtained indicate that the effect of under-sowing velvet beans is generally excellent in soils repeatedly cropped in the previous years, but doubtful in areas that have been fallow for some time.