Two popular critical plane models developed by Fatemi-Socie and Smith-Watson-Topper were derived from the experimental observations of the nucleation and growth of cracks during loading. The Fatemi-Socie critical plane model is applicable for the life prediction of materials for which the dominant failure mechanism is shear crack nucleation and growth, while the Smith-Watson-Topper model, for materials that fail predominantly by crack growth on planes perpendicular to the planes of maximum tensile strain or stress. The two critical plane models have been validated primarily by in-phase and 90° out-of-phase loading, and few, on the complex, non-proportional loading paths. A successful critical plane model should be able to predict both the fatigue life and the dominant failure planes. However, some experimental studies indicate the 304 stainless steel has the two possible failure modes, shear and tensile failure dominant, depending on the loading mode and stress and strain states. Thus a single critical plane model would be impossible to correlate data in life and failure planes in all life regimes, and the valid concept of using the critical plane model is challenged. Since the critical plane approach has not been used to evaluate the experimental fatigue lives of the 304 stainless steel tubular samples under complex, non-proportional loading, it is the objective of our study to examine how the two critical plan models would be correlated to the experimental fatigue data in terms of life and crack growth or failure planes.