Deterministic models assessed the effects that contaminated dung from insecticide-treated cattle had on populations of three hypothetical species of dung fauna that dispersed randomly and could double their numbers every 1–28 weeks at low density. Insecticide was allowed to kill 2–98 % of adults and prevent 16–100% of breeding in pats produced immediately after cattle treatment, with toxicity declining to < 1% in pats produced 2–23 days later. Treatment intervals were 10–40 days. The modelled impact of insecticide was affected little by approximately four-fold variations in: length and density dependence of the attractive life span of pats, frequency of pat occupation by immature adults, distribution of pat toxicity during treatment interval, and changes in dispersal rates due to age and population density. Of greater importance were variations in: pat toxicity, treatment interval, frequency of pat occupation by breeding adults, density dependence of recruitment and death, natural adversity and mortality in dormancy, general rate of dispersal, and the size and shape of the area with treated cattle. Overall, it seemed that wide variations in the impact of contamination will occur in the field, but in many situations the risk to dung fauna can be substantial, especially for slow breeding beetles, and muscoids contacting insecticide on cattle. Risk extends outside the treated areas, for a distance equal to several daily displacements of the insects. Untreated refuges for species survival should be compact blocks at least 25 daily displacements wide.