Amoeba-resistant bacteria (ARB), such as Legionella spp., are currently regarded as potential human pathogens that live in the natural environment, and thus their habitat is regarded as a reservoir of human pathogens. To detect ARB in human and environmental samples, co-culture with amoebae has been demonstrated to be an efficient tool. However, to date, only water samples from cooling towers and hospital water supplies have been investigated as possible reservoirs of ARB using this procedure. In the present study, we studied the ARB population of 11 diverse soil and sand sources in proximity to human environments; these sources included the university, the station, hospitals, the square, parks and public beaches in the city of Marseilles, France. As a result, a total of 33 different species of ARB were identified. The ability to grow within and/or lyse amoebae was demonstrated, for what is believed to be the first time, for several species; moreover, 20 of the isolates (61 %), including Streptococcus pneumoniae, have been described as human pathogens. However, Legionella spp. were not isolated. Four isolates are likely to be the members of new or uncharacterized genera or species, and their capability to be human pathogens needs to be determined. This preliminary work demonstrates that soils and sands in the vicinity of humans are reservoirs of human pathogenic ARB.