Aims: To compare a recently developed monoclonal antibody (MAb) based sandwich ELISA (sELISA) with an immunomagnetic separation (IMS) method for the detection of Escherichia coli O157 in bovine faeces. Methods and Results: Faecal samples from 345 cattle were obtained from eight farms in Northern Ireland, in which human disease due to E. coli O157 had occurred. Both assays detected E. coli O157 on five of the farms and the phage-type of the majority of the bovine strains were the same as the corresponding human isolates. Similar numbers of the organism were detected by the two methods, 59 by the sELISA and 53 by the IMS procedure, 39 of the positive samples being common to both. Twenty samples were sELISA positive/IMS negative. Conclusions: If the IMS is regarded as the gold standard, then the sELISA is less sensitive and less specific, but under the conditions used sELISA positive results were obtained from all positive farms, and the sELISA gave a presumptive positive a day earlier than the IMS method. Significance and Impact of the Study: The sELISA has the potential to be used as a rapid method for screening large numbers of samples for E. coli O157, but further work is required to determine its specificity.