Abstract A quantitative microbial risk assessment incorporating Monte Carlo simulations is described which estimates the probability of Escherichia coli O157 infection of humans by visiting pasture previously grazed by cattle. The risk assessment is performed for a number of scenarios including a variation in the grazing period prior to the human visit, the duration of visit (8-h day or 24-h camp) and the level of E. coli O157 shed by the cattle. Assuming the cattle have been on the field for 28 days, followed directly by a human visit, and the proportion of animals shedding the organism are as described in previous surveys 5±1% (Synge, B.A., Gunn, G.J., Ternent, H.E., Hopkins, G.F., Thomson-Carter, F., Foster, G., Chase-Topping, M., McKendrick, I., 2001). Prevalence and factors affecting the shedding of verocytotoxin producing Escherichia coli O157 in beef cattle in Scotland. In: Concerted Action CT98-3935 Veroctotoxigenic E. coli in Europe, 5. Epidemiology of Verocytotoxigenic E. coli, Dublin, pp. 98–103.), a probability of infection of 0.1% is attained for 8- and 24-h periods when the cattle are shedding approximately 10 3 and 10 4 CFU g −1, respectively. Monte Carlo simulations demonstrated that risk mitigation strategies of removing cattle from the pasture 4 weeks prior to the human visit in addition to physical removal of faeces showed significant reductions in potential infection rates.