Reductions in the yields of spring wheat (cv Bastion) caused by interference from spring oat (cv Selma), considered as a weed, were estimated in field studies on post-emergence weed control in Dijon (France). The growth and development of both plants in competition was different for each of the 3 experimental years. Wheat yield losses were significantly greater on oat-infested plots than on plots treated with diclofop-methyl. A hyperbolic model provided an excellent fit to data for both wheat-grain yield and ear density as a function of oat density, measured at the beginning of tillering. On infested plots where the oats was left to grow until harvest, the number of grains per wheat ear decreased with increasing oat density, and weight-per-grain was not affected. On plots where diclofop-methyl was not applied, chemical control of broad-leaf weeds did not increase wheat yield. On plots controlled with diclofop-methyl, the per cent wheat yield losses differed with the chemical weed control of broad leaf weeds used. When diclofop-methyl and 2,4-D were applied on the same day, the number of wheat ears m-2 was lower than for other diclofopmethyl + antidicotyledoneous herbicide applications. These results show that competitive interference between weeds and crop plants can be modified by the sequence of herbicides applied, so altering the measured crop yield components.