Recent studies have shown that Escherichia coli can survive in different environments, including soils, and they can maintain populations in sterile soil for a long period of time. This indicates that growth-supporting nutrients are available; however, when grown in non-sterile soils, populations decline, suggesting that other biological factors play a role in controlling E. coli populations in soil. Free-living protozoa can affect the bacterial population by grazing. We hypothesized that E. coli strains capable of surviving in non-sterile soil possess mechanisms to protect themselves from amoeba predation. We determined the grazing rate of E. coli pasture isolates by using Dictyostelium discoideum. Bacterial suspensions applied to lactose agar as lines were allowed to grow for 24 h, when 4 µL of D. discoideum culture was inoculated in the center of each bacterial line. Grazing distances were measured after 4 days. The genomes of five grazing-susceptible and five grazing-resistant isolates were sequenced and compared. Grazing distance varied among isolates, which indicated that some E. coli are more susceptible to grazing by protozoa than others. When presented with a choice between grazing-susceptible and grazing-resistant isolates, D. discoideum grazed only on the susceptible strain. Grazing susceptibility phenotype did not align with the phylogroup, with both B1 and E strains found in both grazing groups. They also did not align by core genome phylogeny. Whole genome comparisons revealed that the five most highly grazed strains had 389 shared genes not found in the five least grazed strains. Conversely, the five least grazed strains shared 130 unique genes. The results indicate that long-term persistence of E. coli in soil is due at least in part to resistance to grazing by soil amoeba.