A study was conducted over 3 years (1998-2000) to investigate larval availability of gastrointestinal nematodes from faeces of cattle reared under different parasite control schemes. These cattle were part of a parallel, but separate grazing trial, and were used as donor animals for the faecal material used in this experiment. At monthly intervals, faeces were collected and pooled from three groups of first-season grazing cattle. These groups were either untreated, ivermectin bolus treated or fed the nematophagous fungus Duddingtonia flagrans. The untreated and fungus treated animals were infected with gastrointestinal nematodes and the number of eggs per gram (epg) pooled faeces ranged between 50 and 700 in the untreated group and between 25 and 525 epg in the fungus treated group. Each year between June and September, artificial 1 kg dung pats were prepared and deposited on pasture and protected from birds. The same treatments, deposition times and locations were repeated throughout the study. Larval recovery from herbage of an entire circular area surrounding the dung pats was made in a sequential fashion. This was achieved by clipping samples in replicate 1/4 sectors around the dung pats 4, 6, 8 and 10 weeks after deposition. In addition, coinciding with the usual time of livestock turn-out in early May of the following year, grass samples were taken from a circular area centred where the dung pats had been located to estimate the number of overwintered larvae, which had not been harvested during the intensive grass sampling the previous year. It was found that recovery and number of infective larvae varied considerably within and between seasons. Although the faecal egg counts in 1999 never exceeded 300 epg of the faecal pats derived from the untreated animals, the abnormally dry conditions of this year generated the highest level of overwintered larvae found on herbage in early May 2000, for the 3 years of the study. Overall, biological control with D. flagrans significantly reduced larval availability on herbage, both during and between the grazing seasons, when compared with the untreated control. However, the fungus did not significantly reduce overwintered larvae derived from early season depositions (June and July), particularly when dung pats disappeared within 2 weeks after deposition. Very low number of larvae (<3 per kg dry herbage) were sporadically recovered from grass samples surrounding the ivermectin bolus faecal pats.