Field trials at Beverley (19911, Salmon Gums (1991; 2 sites) and Merredin (1992; 2 sites), each with 5 rates of nitrogen (N) and 3 levels of weed control, were used to investigate the effect of weeds and N on wheat grain yield and protein concentration during 1991 and 1992. Weeds in the study were grasses (G) and broadleaf (BL). Weeds reduced both vegetative dry matter yield and grain yield of wheat at all sites except for dry matter at Merredin (BL). Nitrogen fertiliser increased wheat dry matter yield at all sites. Nitrogen increased wheat grain yield at Beverley and Merredin (BL), but decreased yield at both Salmon Gums sites in 1991. Nitrogen fertiliser increased grain protein concentration at all 5 sites-at all rates for 3 sites [Salmon Gums (G) and (BL) and Merredin (G)] and at rates of 69 kg N/ha or more at the other 2 sites [Beverley and Merredin (BL)]. However, the effect of weeds on grain protein varied across sites. At Merredin (G) protein concentration was higher where there was no weed control, possibly due to competition for soil moisture by the greater weed burden. At Salmon Gums (G), grain protein concentration was greater when weeds were controlled than in the presence of weeds, probably due to competition for N between crop and weeds. In the other 3 trials, there was no effect of weeds on grain protein. The effect of weeds on grain protein appears complex and depends on competition between crop and weeds for N and for water at the end of the season, and the interaction between the two.