Experiments were conducted under controlled conditions to test whether the residues of cover crops or alley crops were phytotoxic to vegetable seedlings in a no-tillage cropping system. Soil and residue samples were collected from a field experiment where subterranean cloverTrifolium subterrcmeumL. var.brachycalycinumand annual ryegrassLolium rigidumGand. were grown as cover crops, and lucerneMedicago sativaL. was grown as an alley crop. Chemically inert black and white polyethylene beads were used as control treatments and provided a measure of the physical mulch effect in the absence of phytotoxic leachates from plant residues. The seedling growth of lettuceLactuca sativa, broccoliBrassica oleraceavar.italicaand tomatoLycopersicon esculentum, was reduced by ryegrass and clover cover-crops. By eliminating factors such as nutrients, soil structure and soil pathogens, we inferred that phytotoxic leachates from the cover crops were responsible for the reduced plant growth. The phytotoxic period of clover ended after 8 weeks in the field and 6 weeks in the laboratory, when conditions for the decomposition of phytotoxic compounds were favourable. The phytotoxic effect of ryegrass lasted longer, and was more severe than clover. Lucerne tops, which are used as a mulch in a vegetable alley-cropping system, caused no growth reduction. Whereas the mulches did improve plant growth by the modification of soil water and temperature regimes, some or all of these benefits were eroded by phytotoxic leachates from the cover crop residues. In contrast to cover cropping, there were no problems relating to phytotoxicity with using undecomposed lucerne shoots as a mulch in an alley-cropping system.