Changing agricultural practices from conventional to conservation tillage generally leads to increased weed populations and herbicide use. To gain information about the possible use of lucerne (Medicago sativa L.) cover crop as an alternative and sustainable weed-control strategy for winter wheat (Triticum aestivum L.), an experiment was performed at Thiverval-Grignon, France, from 2008 to 2010. We compared conventional and reduced tillage as well as the presence and absence of living mulch (i.e. lucerne) on weeds and wheat production. Percentage soil coverage and aboveground biomass of wheat, lucerne and weeds were measured at the end of grain filling. Weed communities were analysed in terms of composition and diversity. During both seasons, wheat biomass did not significantly decrease in reduced-till trials compared with conventional ones (7.0 and 7.2 t ha^–1, respectively, in 2008–09; 6.9 and 7.1 t ha^–1 in 2009–10). Regardless of soil management, the percentage soil coverage by wheat significantly decreased when it was intercropped, although wheat biomass was not significantly reduced compared with the sole crop. To minimise cash-crop losses, we studied the competition between wheat, lucerne and weeds, testing various herbicide strategies. Early control of lucerne allowed better balance between weed control and wheat development. In addition, weed communities varied among treatments in terms of abundance and composition, being reduced but more varied in plots associated with lucerne. A functional group analysis showed that grasses benefited from reduced-till conditions, whereas problematic weeds such as annuals with creeping and climbing morphologies were substantially reduced. In addition, annual and perennial broad-leaf species with rosette morphology were also significantly decreased when lucerne was used as living mulch. Wheat production in reduced-till conditions intercropped with lucerne living mulch may be useful for integrated weed management, reducing the need for herbicides.