Escherichia coli is a natural inhabitant of the gut and E. coli levels in water are considered internationally to be an indication of faecal contamination. Although not usually pathogenic, E. coli has been linked to numerous foodborne disease outbreaks, especially those associated with fresh produce. One of the most common ways through which E. coli can be transferred onto fresh produce is if contaminated water is used for irrigation. In this study, a total of 81 confirmed E. coli strains were isolated from the Plankenburg River as part of three separate studies over 3 years. During sampling, E. coli levels in the river were above the accepted levels set by the World Health Organization and the South African Department of Water Affairs and Forestry for safe irrigation of fresh produce, which indicates that transfer of E. coli during irrigation is highly probable. Multiplex polymerase chain reaction screening for pathogenic gene sequences revealed one enteroaggregative positive strain and four enteropathogenic positive strains. The four enteropathogenic strains were also found to be resistant to three or more critically and highly important antibiotics and were therefore classified as multidrug resistant strains. These results show that E. coli with enteropathogenic potential and multiple antimicrobial resistance properties has persisted over time in the Plankenburg River. Although the presence of several potential pathogens in the Plankenburg River has been reported, 8 the occurrence and types of pathogenic E. coli are not known. The aim of this study was therefore to determine the number, types and antibiotic susceptibility of potential pathogenic E. coli present in the Plankenburg River. Materials and methods Sampling Water samples from which E. coli was isolated were collected from the Plankenburg River (Stellenbosch, Western Cape Province) at three sites (P-0, P-1 and P-2). 8 The P-0 sampling site was about 5 km upstream of the P-1 site and was selected specifically to assess the impact of an informal settlement and industrial area on the water quality of the Plankenburg River. However, for large parts of each year, especially during the dry summer months, this site had no flowing water and therefore it was not sampled at the same frequency as the other sites. Two of the sites-P-1 and P-2-are situated downstream of an informal settlement and industrial area and have shown high levels of faecal contamination. 8 River water sampling was done in accordance with the SANS 5667-6 method 9 ; all samples were 1 L in volume.