Six groups of four ewes with twin lambs grazed six paddocks during two consecutive seasons from May to September in a two-factorial study on stocking rate (9, 13 and 17 ewes ha-1) and infected (designated I-low, I-medium and I-high) versus clean pasture (C-low, C-medium and C-high). All animals were treated with anthelmintic at turn-out and the C-groups were further treated fortnightly. The two sets of paddocks received 0, 80 and 160 kg N fertilizer ha-1 annually with increasing stocking rate. Ewes were withdrawn at the end of July, and all lambs were slaughtered at the end of the grazing season for total worm counts. Nematode infections were subclinical during the first year. The second year, diarrhoea and soiled hindquarters were present from early August onwards in all I-groups with highest incidence at high stocking rate. The faecal egg counts of I-ewes were similar the first year but were substantially higher in I-medium and I-high compared with I-low in the second year. The lambs had differences in faecal egg counts related to stocking rate both years (P < 0.05) and group I-high had faecal egg counts ten-fold higher than I-medium and I-low at the end of the second year. A significant effect of the interaction of stocking rate times +/- infection was found in serum-pepsinogen levels in lambs (P < 0.05). Worm burdens (predominantly Trichostrongylus vitrinus, Ostertagia circumcincta and Nematodirus filicollis) were not related to stocking rate in lambs at the end of the first year whereas there was a significant effect of stocking rate on T. vitrinus burdens the second year. Pasture larval counts and results of tracer lambs indicated a positive relationship between stocking rate and pasture infectivity. It is concluded that the level of ovine nematode infections, especially Trichostrongylus spp., in weaned lambs increased with increasing stocking rate even though pasture production was enhanced by increased rates of fertilizer application.