Thermal full‐thickness burns involving 2 per cent of the body surface area were produced in guinea pigs. DNase‐ and coagulase‐negative staphylococci cultured from the animals' normal bacterial flora were used as interference strain. They were applied to the burns of the test guinea pigs. As controls, guinea pigs were used, which were burned in the same way but not inoculated with the interference strain. An easily identifiable strain ofStaphylococcus aureuswas subsequently sprayed on the burns in some of the series. In the burns which were not sprayed with this test strain, records were kept only of spontaneously occurringStaphylococcus aureus.In all the series the animals in the test groups showed significantly less growth ofStaphylococcus aureusboth with respect to the percentage of positive animals and to the colony counts from these animals.Staphylococcus aureuspositive guinea pigs thus were found in only 27 per cent in the test groups, but in 73 per cent of the controls. From the number of colonies per culture plate “growth scores” were calculated from 0 to 4 according to an arbitrary scale. The mean growth scores per animal ofStaphylococcus aureuswithin the first two weeks after burning was 3.1 in the test guinea pigs and 8.8 in the controls. Possible practical consequenses of the observations made in the experiments are discussed, as for instance the possibility of bacteria belonging to the indigenous flora or other bacteria of lower pathogenicity to act as interference strains against hospital strains of pathogenic stap