The invention discloses a method for determining fracture toughness and strength by using double-cut small-sized specimens. The method comprises the following steps: pouring the specimens, cutting out cracks, loading a static force, recording the peak loads, calculating the nominal strengths sigman, calculating geometric parameters A of the specimens, and finally performing regression analysis to obtain the fracture toughness K and the tensile strength f of a material without a size effect. Based on an idea of progressive extrapolation as well as theoretical analysis in elastoplastic fracture mechanics, material parameters of infinitely large slab specimens (the material parameters of the infinitely large slabs have no size effect) are extrapolated from experimental data of finite-size specimens, so that the fracture toughness and the strength of the material without the size effect can be determined only according to the key loads F of the small-sized specimens; strict current domestic and foreign specifications on the sizes and the types of the specimens, loading conditions and the like are not required to be met, and the stability of the double-sided tensile specimens is better than that of the single-sided tensile specimens, so that operation is easier.