In an ideal pasture, sown species would persist indefinitely and there would be no weeds. 'Poor persistence' occurs when the desirable species are replaced by weedy, undesirable species. Unfortunately, poor persistence of sown pasture plants is the reality that most farmers experience. In this paper we review the literature on three key factors leading to poor persistence and weed ingress: 1) inter-specific competition; 2) sward disturbance; and 3) propagule pressure. We reveal that grazing and other management decisions, pasture insect pests, nutrients, climate and the composition of the sown species mix have a critical influence on these factors, and hence over the replacement of sown pasture plants by weeds. The key practical message from this review is that farm practices that improve the vigour of sown species, reduce pasture gaps and minimise weed propagule pressure will improve the persistence of sown species and minimise weed ingress. An outstanding deficiency is our lack of knowledge of the economic impacts, and hence weed status, of most of the ca 187 volunteer plant species that may invade and replace sown plants in New Zealand's pastures.