Fatigue performance is crucial for gas turbine components, and it is greatly affected by the manufacturing processes. Ability to predict the expected fatigue life of a component based on surface integrity has been the objective in this work, enabling new processing methods. Alloy 718 samples were prepared by different machining setups, evaluated in fatigue testing and surface integrity investigations. These results generated two predictive statistical multi-variate regression models. The fatigue correlated well with roughness, residual stresses and deformation. The two models showed great potential, which encourages further exploration to fine-tune the procedure for the particular case.