Although cattle are reservoirs, no validated method exists to monitor Shiga toxin-producing Escherichia coli O157 (STEC O157) on farms. In 29 Midwestern United States feedlot pens we compared culturing faeces from the individual cattle to: (1) culturing rope devices that cattle rub or chew; and (2) culturing a composite of faecal pats. Eighty-six per cent (68-96%) of pens were classified correctly using rope devices to detect pens with at least 16% of the cattle shedding STEC O157 [sensitivity=82% (57-96%); specificity=92% (62-100%)]. Ninety per cent of pens (73-98%) were classified correctly using composite faeces to detect pens with at least 37% of the cattle shedding STEC O157 [sensitivity=86% (42-100%); specificity=91% (71-99%)]. Ranking pens into three risk levels based on parallel interpretation of the pen-test results correlated (Spearman's r=0.76, P<0.0001) with the pen's prevalence. This strategy could identify pens of cattle posing a higher risk to food safety.