[en] Creep-resistant austenitic and ferritic/martensitic steels are used for pressure vessels, reactor internals and piping applications in nuclear power plants. Creep-fatigue is one of the dominant failure modes in these materials during the lifetime of a reactor. In this study, a modeling approach for the creep-fatigue crack growth is presented. The approach combines a constitutive model for creep with a strip-yield model for crack growth. In the constitutive model, creep deformation results from the dislocation and dipole generation and annihilation within the microstructure. Crack growth is modeled using a strip-yield methodology. In this approach, the crack plane is meshed with one-dimensional bar elements. The deformation and fracture of the elements located in the crack plane and ahead of the crack tip are described using the constitutive model. The crack advance results from the progressive failure of the crack plane elements. Simulations of creep-fatigue crack growth in compact-tension specimens of Grade 91 (modified 9Cr-1Mo) steel are presented. Several ratios of creep and fatigue during a loading cycle at different temperatures are considered. The advantage of this modeling approach is that it offers a method to extrapolate models built using accelerated creep-fatigue testing data to more realistic service loading scenarios. (author)