Dry snow slab avalanche release is preceded by two fracture mechanical processes: shear failure of a weak layer or an interface within the snowpack, followed by tensile failure of the overlaying slab. For a slope stability analysis based on fracture mechanics, the fracture toughness of snow has to be known. The purpose of this work was to evaluate snow fracture toughness in mode 1, to determine to what extent it is affected by the specimen size and shape and to search for adequate correction methods. Edge-cracked beam-shaped snow specimens cut from homogeneous layers of naturally deposited snow were subjected to three-point bending and cantilever beam tests. To describe the size dependence an empirical size effect law and the FAD (failure assessment diagram) approach were explored. By comparing the three-point bending with the cantilever beam tests a shape dependence of the toughness was found. The fracture process zone was estimated to be in the order of at least one centimetre. Due to the large size of the fracture process zone a dependence of the toughness on the specimen size has to be expected, as it is typical for non-linear, quasi-brittle materials. Experiments with four different specimen sizes clearly confirmed that toughness is size dependent, possibly up to the scale of a slab avalanche. Preliminary results suggest that the actual fracture toughness might be twice as large as the one determined experimentally. Therefore, size correction functions will be essential to transform toughness data of laboratory-scaled experiments to the scale relevant for snow slope stability models.