This work focuses on using representative corrosion damage morphology, determined from fractography, to predict crack propagation behavior in a test article under fatigue load. A dogbone specimen (AA7075-T651) with a simulated fastener hole was pre-corroded using atmospheric corrosion conditions (RH = 80%, salt load = 1000 mg/m², T = 25^o C). A CFRP insert was placed in a through hole on the side surface of the dogbone to represent a fastener system and to promote local corrosion through galvanic coupling. A thin non-conductive separator was placed between the CFRP insert and the AA7075-T651 through hole to prevent physical contact between those two materials. The specimen was salt dosed and placed in an environmental chamber with controlled relative humidity and temperature. A zero-resistance ammeter test of the CFRP/AA7075-T651 couple was performed for 10 days to induce galvanic corrosion damage on the surface surrounding the through hole mouth. Following pre-corrosion, the dogbone specimen was tested under constant amplitude fatigue load (S_max = 25.11 ksi, R = 0.89) in laboratory air; the total fatigue life measured experimentally was 230,135 cycles. Following fatigue testing the failed specimen was inspected for corrosion-related crack initiation sites and final crack shape using SEM imaging. A computational fracture mechanics model, based on the boundary integral method, was developed to simulate the fatigue test. Fractographic imaging was used to size and locate initial flaws in the through hole. To approximate the anticipated short-to-long crack transition, the Hartman–Schijve equation was fitted to a crack growth kinetic data set and used to drive the simulation model. A range of model inputs were selected to understand the sensitivity of fatigue life predictions to the fracture mechanics model inputs used to represent the complex corrosion damage morphology. The computational model produced results that were in reasonable agreement with the experimental test.