The experiment pretended to watch the soybean's behavior submited to pruness by realizing fertilization with boron, cinc and sulphur and their influence over seed and f or age yield. Experimental design used was blocks by random completely, with five repetitions and nine treatments set up by: an absolute control (no pruned, no fertilized), a normal control (pruned, no fertilized), other treatments were pruned and fertilized by this way: boron, cinc, sulphur, boron + cinc, boron + sulphur, cinc + sulphur, boron + cinc + sulphur, this fertilization was made to soil in sowing moment. The variety was Andree-23 (600.000 plants /ha.), with forrows separated by 40 cm, cutting was made 38 days after emergetion. The whole pruness makes a negative effect over seed yield by decreasing it about 53 percent, but arise branche number. Fertilization only was good for height plant and empty pods reduction; over other variables (pod number/plant, seed number/pod, a hundred seed average weight, etc.) wasn't evident any incidence.