A trial compared three types of seed treatments using five different insecticides: trichlorfon, isophenphos, carbofuran, methomyl, and omethoate. Coating method had no effect on insecticidal activity, although seedling growth and subsequent pasture production were greater from ryegrass seed loaded with fertiliser than from experimental or commercial ('Prillcote') dressings incorporating insecticide only. Only isophenphos and carbofuran controlled larvae, and the former gave greater pasture production than all other treatments. These results were supported by analyses of insecticide residues which showed isophenphos and carbofuran to be the most persistent.