SummaryPlots sown with a range of test species were treated with chlortoluron, nitrofen, propyzamide or terbutryne applied at several doses either by rotary atomizer (30 1 ha−1), conventional hydraulic nozzle (250 1 ha−1) or dribble bar (with orifices spaced at 2‐cm intervals) (520 1 ha−1). The performance of chlortoluron, propyzamide and terbutryne was similar with all the methods used, regardless of soil and weather conditions. Vertical movement in the soil was similar for each application method. Horizontal variability following dribble‐bar application was measured for chlortoluron and was very large across short distances but weed control was as uniform as with the other application methods.In a further experiment, the phytotoxicity of chlortoluron applied by each of the three methods was less in a seed bed prepared by minimum tillage following straw burning than in a ploughed and cultivated seed bed. Activity following minimum tillage and straw baling was intermediate. This difference in phytotoxicity probably resulted from greater adsorption in the surface layers of the minimal cultivation treatments.