S U M M A R YSix experiments were carried out over 2 years (1985 and 1986) in commercial spring barley crops on calcareous soils in Hampshire. Herbicides increased grain yield by 005-1-21 t/ha, and fungicides by 001-0-68 t/ha. Herbicide use affected each of the yield components ears/plant, grains/ear, 1000-grain weight and harvest index in one or more cases but the effects were not consistent between experiments or consistently related to yield increases. Fungicide use had little effect on yield components other than 1000-grain weight, which was significantly increased in all experiments. There were no significant herbicide x fungicide interactions. Large differences in yield were recorded between field edge (headland) and midfield plots, with headland plots outyielding the midfield in two of the three experiments where this effect was considered.The effects of fungicide use were accounted for by associated increases in the green lamina area of the flag leaf and second leaf at the late milk stage of grain development. This is consistent with previous work. Differences in yield response to herbicide between sites were linearly related both to weed numbers early in the season and to weed biomass at harvest. Weeds did not always directly replace crop biomass; in some experiments the combined dry matter production was greater when herbicide was not used. However, the results indicate that, where a diverse weed flora composed of species of moderate to low competitive ability is present, simple weed counts may provide a useful indication of potential yield loss.It is concluded that headland areas can often be as productive as the rest of the field, but a greater understanding of the factors involved is needed if this potential is to be realized.