The stress field has been determined at eight different sites in France (four in crystalline or metamorphic rocks, four in sedimentary formations) by hydraulic tests in boreholes. For seven of these sites the complete stress field has been determined using the hydraulic tests on pre‐existing fractures (HTPF) inversion method. Validity of the results is demonstrated by the good fit between a priori and a posteriori values for the data and by the low values of the a posteriori standard deviation on the unknowns. For four sites, results have been compared with those derived according to the hydraulic fracturing theory. The two methods yield comparable results for the orientation of the maximum horizontal principal stress σH, a generally satisfactory fit for the magnitude of the minimum horizontal principal stress σh, but a very poor agreement for the magnitude of σH. This latter misfit has been attributed to the effect of fluid percolation prior to the actual opening of the fractures. Below a depth of 500 m, a homogeneous σH direction (N150°E) has been determined at the crystalline sites (Auriat, Echassières, Le Mayet de Montagne, Chassoles), all located in the northern Massif Central. For three of these sites the vertical stress is significantly lower than the weight of overburden as computed from the rock density and the depth of the corresponding measurement. At Auriat and Echassières the stress field is consistent with a mostly strike‐slip faulting regime. At Le Mayet de Montagne and Chassoles the maximum stress is vertical but nearly equal to σH. The stress field has been found to be much more heterogeneous at two of the four sites in sedimentary rocks because of the large variability in mechanical rock properties. In such heterogeneous formations, inversion with the HTPF method must be limited to those data which pertain to the same rock horizon. However, because of its structure this heterogeneity does not restrict the possibility of defining large‐scale uniform principal stress directions. All the results are consistent with the local seismotectonics.
