The problem of physically incorrect results due to neglecting the interaction of the crack edges in fracture mechanics is discussed. The cases of dynamic loading, interface cracks, and thermal loading in fracture mechanics are analyzed. The main conclusion is that the majority of results obtained in these areas of fracture mechanics are physically incorrect Keywords: fracture mechanics, physically incorrect results, interaction of crack edges Introduction. By physically incorrect results in fracture mechanics, we will mean results of mathematical modeling that disregards certain physical factors of fracture phenomena studied within the framework of fracture mechanics. In this connection, physically correct results in fracture mechanics will be meant results of mathematical modeling that takes into account all physical factors of fracture phenomena studied within the framework of fracture mechanics.A number of such factors are quite obvious a priori and should be taken into account in formulating problems of fracture mechanics. At least, neglecting them should be analyzed and commented. However, there are situations where the necessity of allowing for some factors becomes clear a posteriori, after solving a specific problem of fracture mechanics. Examples of such situations will be given below.The present paper analyzes the physical correctness of results in fracture mechanics that are united by the phenomenon of interaction of the edges of crack near its tips or throughout its length. Making consistent and rigorous allowance for the interaction of crack edges is possible only within the framework of nonlinear fracture mechanics where the linear equations of motion and constitutive equations can be used with nonlinear boundary conditions that change during deformation. As termed in [18], these problems of fracture mechanics are nonclassical because Condition 3 [18] that the boundary conditions do not change during deformation is not satisfied.Nowadays, the overwhelming majority of researchers neglect the interaction of crack edges and solve problems of fracture mechanics in a linear formulation. Such an approach assumes, making no clear provision and no comments, that the crack edges as if penetrate each other during deformation, which obviously leads to physically incorrect results. The sole exception is the approach [1-3, 17, 20-22, 30] for dynamic fracture mechanics that is based on a nonlinear formulation and accounts for the interaction of crack edges. As termed above, the results in dynamic fracture mechanics obtained with this approach are physically correct.In what follows, we will analyze, in a brief and simple manner, the physical correctness of results in fracture mechanics in the cases of dynamic loading, interface cracks between two materials, and thermal loading. We will avoid detailing mathematical methods for a wider range of experts to perceive the matter. The list of references does not cover all well-known publications on the research areas being discussed and includes only typical publications.