A B S T R A C TWe present predictions and measurements of fatigue crack growth rates in plastically bent aluminium 2024-T351 beams. Beam bending and fatigue were carefully controlled to minimize factors other than residual stress that could affect the fatigue crack growth rate, such as large plastic strains or residual stress relaxation. The residual stress introduced by bending was characterized by a bending method and by the slitting method, with excellent agreement between the two methods. Crack growth rates were predicted by three linear elastic fracture mechanics (LEFM) superposition based methods and compared to experimental measurements. The prediction that included the effects of partial crack closure correlated with experimental data to within the variability normally observed in fatigue crack growth rate testing of nominally residual stress free material. Therefore, we conclude that crack growth through residual stress fields may be predicted using the concept of superposition as accurately as crack growth through residual stress free material, provided that the residual stress is accurately known, the residual stress remains stable during fatigue, the material properties are not changed by the introduction of residual stress, and that the effect, if any, of partial crack closure is taken into account. a = crack length B = specimen thickness da/dN = fatigue crack growth rate h(x,a) = weight function K A = stress intensity factor due to applied external forces K max = applied stress intensity factor at P max K min = applied stress intensity factor at P min K ssy = applied stress intensity factor at which small scale yield criteria is violated K open = the stress intensity factor due to P open K R = stress intensity factor due to residual stress K T = total stress intensity factor K T max = total stress intensity factor at P max K T min = total stress intensity factor at P min K T ssy = total stress intensity factor at which small scale yield criteria is violated