SummaryThree field experiments were carried out in organically grown winter wheat in Denmark. The treatments were sowing time (normal or late sowing) and false seedbed, row width (12 and 24 cm) and weed control method [untreated; mechanical weed control (weed harrowing at 12 cm supplemented with inter‐row hoeing at 24 cm); and herbicide weed control]. Weed biomass in midsummer was greatest on plots sown at the normal sowing time (compared with delayed sowing) and was reduced by mechanical or chemical weed control (compared with untreated plots). Row width alone had no influence on weed biomass, but in the experiment with high weed pressure, the more intensive mechanical weed control used at a row width of 24 cm reduced weed biomass. Normal sowing time tended to give higher yields, but this was only statistically significant in one of the three experiments. Wide rows gave a yield decrease in the experiment with low weed pressure. The effect of weed control on yield was dependent on the weed pressure. At low weed pressure, mechanical weed control caused a yield decrease compared with untreated or herbicide treated. At intermediate weed levels there were no differences, whereas at high weed pressure, mechanical weed control and herbicide treatment caused a yield increase compared with untreated. False seedbeds were shown to contribute to a decrease in the soil seed reserve.