Corrosion fatigue is a critical degradation mechanism for stainless steel components in light water reactor (LWR) nuclear power plants, where the combined action of cyclic loading and high-temperature water environments significantly accelerates fatigue crack initiation. This work aims to investigate corrosion fatigue in stainless steel reactor piping systems by integrating experimental observations with physics-based modeling. Corrosion fatigue experiments were conducted on 304L stainless steel under simulated boiling water reactor (BWR) and pressurized water reactor (PWR) conditions using hollow specimens [1]. Tests were performed under cyclic loading in pressurized water at 300 °C and 12 MPa. For reference, S–N curves obtained from standard specimens tested in air were compared with corrosion fatigue results from hollow specimens exposed to reactor water environments. Crack initiation was monitored in situ using the direct current potential drop (DCPD) technique. Post-test fracture surface and microstructural analyses were carried out using scanning electron microscopy (SEM) and electron backscatter diffraction (EBSD) to identify crack initiation sites, environmentally assisted damage features and underlying mechanisms. Based on the experimental findings, a continuum damage mechanics (CDM) corrosion model was proposed to describe corrosion-assisted fatigue crack initiation. The model incorporates electrochemical kinetics through Faraday’s law and accounts for mechanically driven oxide film rupture and corrosion activation. The formulation is implemented within a von Mises elastoplastic framework to simulate corrosion pit formation in discontinuous geometries. Armstrong-Frederick kinematic hardening is included to accurately capture the cyclic plasticity. The corrosion damage model is also implemented within a crystal plasticity framework, where Voronoi tessellations are used to represent the polycrystalline grain structure. This approach enables simulation of corrosion damage in uniform geometries such as hollow specimens. Numerical simulations of a representative section of the hollow specimen are performed to estimate crack initiation, providing a mechanistic link between microstructure, environment and fatigue damage accumulation.