This paper evaluates the capability of the theory of critical distances (TCD) coupled with the Fatemi–Socie (FS), Smith–Watson–Topper (SWT), and normal stress range (S‐N) criteria for the lifetime estimation of combined fretting fatigue, that is, the combination of bending and axial loading modes, for 316L stainless steel along with the effect of increasing the grain size of the material. After evaluating the stress field, the point, line, area, and volume methods have been implemented to estimate the lifetimes of fretting fatigue experiments based on the dependency of the critical distance on the number of cycles to failure. Whereas the area and volume methods were overestimating the lifetimes, the point and line methods were more accurate. FS and SWT criteria were proper in lifetime estimation, but the normal stress range (S‐N) criterion was not a safe criterion. At the same time, increasing the grain size of the material leads the TCD to be more conservative.