Peridynamics is a nonlocal numerical framework that enables robust simulation of damage evolution and crack initiation and propagation in solids [1,2,3]. However, peridynamics requires a large number of interacting material points to calculate internal forces for each material point, and fully peridynamic analyses can be computationally prohibitive for large-scale problems. To mitigate the computational cost, various coupling strategies have been proposed to combine peridynamics with the finite element method (FEM), which is relatively computationally efficient [4,5]. For large-scale problems under complex boundary conditions, the crack propagation is unpredictable, and an adaptive coupling strategy is essential to preserve both accuracy and efficiency by updating the peridynamic and FEM subdomains in real time. In this work, we develop an adaptive peridynamics–FEM coupling method building on a previously established coupling framework [5]. The proposed adaptive strategy is first verified using a tensile test. Then, the verified strategy is applied to fracture simulations of a specimen containing an initially inclined crack, with an emphasis on quantifying crack growth behavior as a function of crack geometry. The results demonstrate that the adaptive coupling approach reproduces crack paths with accuracy comparable to a fully peridynamic reference solution. Ongoing work will quantify differences in the onset of crack initiation among pure peridynamics, non-adaptive coupling, and the proposed adaptive coupling, and will extend the framework to a wider range of inclined-crack configurations and more complex geometries.