In the experimental field of Godollo University on brown forest soil, the effect of split nitrogen top dressing was studied on the yield. The applications were carried out at four dates - emergence, tillering, shooting, heading - at three fertilizer levels - 120, 160 and 200 kg/ha - using altogether 17 varieties. In 1984 on the average of varieties, the treatments did not deviate reliably from the control, but within the varieties there were deviations observed. With variety Mv 4 the rate of 0 + 80 + 40 + 40, with variety Baranjka 0 + 80 + 40 + 40 and 0 + 80 + 40 + 0, with variety GK Szeged 0 + 40 + 40 + 40 fertilizer applied at spring - at tillering, shooting, heading - seemed to be significantly better than the control and even also the autumn applied - at emergence, tillering - split fertilization rate of 80 + 80 + 0 + 0. In 1985 their was also no difference among the treatments on the average of the varieties, however, with varieties GK Kincso, Bucsanyi 20, Yubileinaya 50, GK Szeged, GK Minaret the treatment seemed to be effective by rate of 80 + 80 + 0 + 0, in contradiction to split application of 0 + 80 + 40 + 0. In 1986, due to the droughty spring, the treatments did not deviate either from the control or from each other on the average of varieties, but within the varieties, depending on the split use there were positive applications, namely with variety Mv 10 the rate of 40 + 80 + 0 + 0, while with GK Kincso and Mv 4 the rate of 80 + 80 + 0 + 0. Due to the great year-effect, there were considerable interactions, so evaluable results were obtained only during the period of 1985-1987 in case of five treatments and three varieties. There was a significant difference between the control and the treatments, but there was not among the treatments. Out of the varieties provably the most superior was the variety Mv 4, this was followed by GK Kincso and finally by GK Boglar.