Broad-leaved weeds have generally been dismissed as less competitive than grass weeds in oilseed rape and therefore their competitive effect on the crop has been ignored. This has limited the development of cost-effective strategies for their control. A survey identified the main weeds of winter oilseed rape in north-east Scotland as chickweed, annual meadowgrass, volunteer barley, mayweed spp., field pansy and shepherd's purse. Of these, only the effect of volunteer barley has been previously researched. Field trials over five years showed that the effects of broad-leaved weeds on the growth and yield of oilseed rape varied between seasons, but clearly demonstrated that these weeds can be very competitive and can cause substantial yield losses. The start of crop stem extension in spring was identified as the critical time for broad-leaved weed competition in oilseed rape. Unless the crop is free of weed competition at this time or is sufficiently vigorous to outgrow weeds, yield losses will result. However, autumn proved to be the optimum time to control broad-leaved weeds and early treatments were found to be most efficacious. The variation in crop response between seasons was largely attributable to seasonal weather and its effects on crop vigour and weed growth. Seasonal weather also influenced the efficacy of reduced doses of herbicides. Metazachlor gave good weed control at reduced doses in vigorous crops, but not in less vigorous crops in poor growing seasons. This variability makes it difficult to predict crops responses to weed competition and the justification for herbicides. Thus a semi-prophylactic approach is proposed, based on a reduced dose of metazachlor applied early with the option of a reduced dose follow-up post-emergence treatment in autumn, if necessary. This is a practical approach which could be adopted on many farms and would benefit the farmer in terms of improved cost-effectiveness and go some way towards satisfying the demand for reduced pesticide inputs.