In a three year study (1994-96) reduced soil tillage variants proved satisfactory for winter wheat grown on Humogley soil in Eastern Croatia in the absence of herbicides. During the experiment the conventional tillage (ploughing, disk-harrowing, standard sowing) did not give significantly higher yields compared to the reduced tillage treatments (disk-harrowing ; tillage by multitiller with chisels ; seedbed preparation (rototiller) + sowing by rotosem ; and ploughing, seedbed preparation + sowing by rotosem). Production of weed biomass at the conventional tillage equalled the treatment tillage by rototiller + sowing by rotosem, but was significantly lower than in other tillage treatments. Totally, nitrogen fertilisation had a significant influence upon wheat yield increase at each N level (140, 170, and 200 kg/ha), whereas the efficacy of fertilisation upon yields was higher at reduced tillage treatments without ploughing than at the conventional tillage. The correlation between crop yield and weed biomass decreased along with the increase of N level. Winter wheat had a higher N optimum than its weed component, which points out that for this crop N fertilisation at reduced tillage systems can be used as a measure for weed control. The reduced tillage variant with rototiller + sowing by rotosem turned out to be entirely competitive with the conventional tillage regarding yield and weed biomass, as well as a higher resistance to weather conditions.