ABSTRACT During the period from September 1996 through November 1996, 10 Dutch dairy farms were visited to collect fecal samples from all cattle present. The samples were examined for the presence of verocytotoxin (VT)-producing Escherichia coli (VTEC) of serogroup O157 (O157 VTEC) by immunomagnetic separation following selective enrichment. Cattle on 7 of the 10 dairy farms tested positive for O157 VTEC, with the proportion of cattle infected varying from 0.8 to 22.4%. On the seven farms positive for O157 VTEC, the excretion rate was highest in calves ages 4 to 12 months (21.2%). In a follow-up study, two O157 VTEC-positive farms and two O157 VTEC-negative farms identified in the prevalence study were revisited five times at intervals of approximately 3 months. Cattle on each farm tested positive at least once. The proportion of cattle infected varied from 0 to 61.0%. Excretion rates peaked in summer and were lowest in winter. Again, the highest prevalence was observed in calves ages 4 to 12 months (11.8%). O157 VTEC strains were also isolated from fecal samples from horses, ponies, and sheep and from milk filters and stable flies. O157 VTEC isolates were characterized by VT production and type, the presence of the E. coli attaching-and-effacing gene, phage type, and pulsed-field gel electrophoretic genotype. No overlapping strain types were identified among isolates from different farms except one. The predominance of a single type at each sampling suggests that horizontal transmission is an important factor in dissemination of O157 VTEC within a farm. The presence of more than one strain type, both simultaneously and over time, suggests that there was more than one source of O157 VTEC on the farms. Furthermore, this study demonstrated that the O157 VTEC status of a farm cannot be ascertained from a single visit testing a small number of cattle.