Soil samples from 76 sites in south-eastern Poland were collected and the presence of entomopathogenic fungi (EPF) was evaluated by the bait larvae method. EPF were detected on the average in 43.5% of samples. The soil usage method (forest, agricultural crop, garden, seminatural habitat), the date of sampling (spring, summer, autumn) and physico-chemical properties of the soil significantly differentiated the total mortality of larvae and the occurrence of EPF and their species composition. Soils of seminatural habitats and agricultural crops had significantly higher level of total mortality of larvae compared to the soils of gardens and forests. The total mortality of larvae in the soil was the highest at autumn. The share of EPF in soils due to the usage methods can be determined in a decreasing sequence: garden, agricultural crops, seminatural habitats, forest. In the soils of agricultural crops I. fumosorosea dominated and in the soils of gardens, forests and seminatural habitats B. bassiana. The share of M. anispoliae sensu lato in soils was rather similar. In the samples obtained in the summer, EPF dominate, and in the samples collected in spring and autumn, other factors dominated mortality of larvae. The species composition of EPF in the soil collected in spring and summer was similar. In the samples taken in autumn, B. bassiana species dominated.