In 2003 the effect of tillage timing (mouldboard ploughing in autumn or spring) and seeding time on weed density in sugar beet was examined in an on-farm trial near Göttingen. Weed density was highest (110 plants m-2) when sugar beets were sown early (7th April). Ploughing in autumn and late sowing (22nd April) led to the lowest weed density (15 plants m-2). 35 plants m-2 were found after ploughing in spring with late sowing. After early sowing, the weed population was dominated by Matricaria spp., while after late sowing Chenopodium album L. was the dominating species. Besides, late sowing resulted in yield losses compared to early sowing. In 2004 weed density was very low (Ø 15 plants m-2) and neither ploughing time nor harrowing pre- or post-emergence had an effect. This was due to the preceding grass-legume mixture and a very dry spring. Weed population was dominated (80-90 %) by C. album L. and volunteer plants from the preceding crop. Harrowing post emergence tended to reduce sugar beet density but still exceeded 80,000 plants ha-1.