A series of experiments was undertaken to determine the effect of different grass species on gastrointestinal nematode parasitism and performance of lambs and the effect of a broader range of herbage species on nematode larval population dynamics. All of the experiments were undertaken at AgResearch Flock House, located in the southern North Island of New Zealand. In the first of two unreplicated grazing experiments, four grass species browntop (Agrostis capillaris cv Muster), tall fescue (Festuca arundinacea cv Au Triumph), Yorkshire fog (Holcus lanatus cv Massey Basyn) and perennial ryegrass (Lolium perenne cv Nui) were compared in single species swards grazed by weaned lambs in each of two years (1991/92 and 1992/93). Swards were grazed to a target sward height of 5 cm by altering stock numbers. On each grass, one third of the lambs were suppressively drenched fortnightly (SD) and two thirds were trigger drenched (TD), when mean faecal egg count on any treatment reached 1500 eggs per gram (epg) in 1991/92 and 1000 epg in 1992/93. In both years, lamb faecal egg counts were higher (P<0.05) in lambs which grazed browntop and tall fescue than in lambs which grazed ryegrass or Yorkshire fog. Parasitism, as measured by tracer lamb nematode burdens, was highest in lambs which grazed browntop, lowest in lambs which grazed ryegrass and Yorkshire fog, and intermediate in lambs which grazed tall fescue (P<0.05). In 1991/92, production losses due to parasitism as measured by the difference in liveweight between SD and TD lambs were higher (P<0.05) in lambs which grazed browntop, tall fescue or ryegrass than in lambs which grazed Yorkshire fog. This pattern was not repeated in the second year. In the second grazing trial, undertaken in 1992/93, lambs grazed tall fescue or Yorkshire fog swards to target heights of 3, 5, or 8 cm. On the tall fescue swards, decreasing sward height increased (P<0.05) tracer lamb nematode burdens, but this was not observed on the Yorkshire fog swards. Also, on the tall fescue swards, there was a significant (P<0.05) production loss associated with parasitism (as measured by liveweight differences between SD and TD lambs), but such a pattern was not observed on the Yorkshire fog swards. In a comparison of the recovery of Trichostrongylus colubriformis larvae from a range of herbages using the modified Baermann technique, greatest numbers were recovered from cocksfoot (Dactylis glomerata cv Wana) and chicory (Chicorum intybus cv Puna), lowest numbers from prairie grass (Bromus willdenowii cv Matua), perennial ryegrass (Lolium perenne cv Nui), and Yorkshire fog (Holcus lanatus cv Massey Basyn), and intermediate numbers from browntop (Agrostis capillaris cv Muster), tall fescue (Festuca arundinacea cv Au Triumph), and white clover (Trifolium repens cv Huia) (P<0.05). There was a greater than two-fold difference in the number of larvae recovered between chicory, which had the highest number of larvae recovered from it, and prairie grass, which had the lowest. In a series of experiments undertaken outdoors, faeces containing known numbers of Ostertagia circumcincta and Trichostrongylus colubriformis eggs were deposited on mini-swards of a range of herbage species, browntop, chicory, cocksfoot, tall fescue, lucerne (Medicago sativa cv Otaio), ryegrass, prairie grass, white clover, and Yorkshire fog. Larvae were recovered from four strata (0-2.5, 2.5-5, 5-7.5, and >7.5 cm above the soil surface) at 2, 4, 6, 8, 11 and 14 weeks after the faeces was deposited on the herbage. These "contaminations" were carried out four times in 1992/93 and 1993/94. Larval development success, defined as the maximum number of larvae recovered on herbage after contamination, differed significantly (P<0.05) between herbage species, being greatest on Yorkshire fog and ryegrass, least on white clover and lucerne and intermediate on the other herbages. The proportion of larvae recovered from the bottom stratum, an inverse measure of the ability of the larvae to migrate vertically, differed (P<0.05) between herbages. It was greatest on Yorkshire fog and prairie grass, least on white clover, ryegrass and browntop with the other herbages intermediate. Larval survival, as estimated by the decline in larval numbers on the herbage, did not differ (P>0.05) between herbages. Two experiments to compare larval development success and migration were done in a glasshouse with mini-swards established in 20 cm diameter plant pots. Four grass species, ryegrass, tall fescue, Yorkshire fog, and browntop were compared. Faeces containing known numbers of Ostertagia circumcincta and Trichostrongylus colubriformis eggs were deposited on swards after cutting to one cm, and the larvae recovered from the four strata (0 - 2.5 cm, 2.5 - 5 cm, 5 - 7.5 cm and >7.5 cm) 4 weeks later. Larval development success did not differ (P>0.05) between grasses. However, the vertical migration patterns were similar to those observed in the outdoor larval dynamics experiments, with larvae concentrated in the bottom stratum of Yorkshire fog but more evenly spread over the four strata in the other grasses. The results from these trials show that, under New Zealand conditions, pasture species can have marked effects on larval development success and larval migration on herbage. This translated into differences in lamb parasitism between grass species. Combining the results from the studies in this thesis with other published results suggests that differences in lamb parasitism between herbage species may vary depending on whether a continuous or discontinuous grazing strategy is used. The studies also demonstrate that on Yorkshire fog swards production losses due to parasitism were lower than for other grasses. It is suggested that parasite levels in lambs which grazed this species were restricted either by physical means through restricted larval migration on herbage or through biochemical means by limiting larval establishment in the gastrointestinal tract of grazing lambs.