In 1969 an experiment with dry beans was started at the Estação Experimental de Jundiaí (State of São Paulo) with the scope to study the effects of the incorporation in the soil of undecomposed vegetative matter, which sources were leguminous plants (Glycine wightii Verdc.) and grass (Melinis minutiflora Béauv.), as well as of the partial shading of the plants. Soil temperatures were determined at depths of two and ten centimeters in all treatments. Air temperature was determined in plots with and without shading. The treatments which consisted in the incorporation of undecomposed vegetative matter at all times produced higher yields than those with shading and control. There was no significant difference in grain yields of treatments with leguminous plants and grass, as well as between shading and control. With incorporation of vegetative matter the variation of the temperature in the soil was a little lower than with shading or control.