The mean stress sensitivity of an imperfective and a near-defect free aluminium cast alloy is analysed in this work. Thereby, metallographical, quasi-static, high cycle fatigue and crack propagation investigations are conducted. The specimens are taken out from an Al–Si–Cu sand cast component with either T6 heat treatment or hot isostatic pressing (HIP) condition. As the specimens are extracted from similar positions, technological size effects in terms of varying microstructural properties, such as the dendrite arm spacing and micro pore size distribution, are suppressed. The crack propagation tests enable the build up of load stress ratio dependent crack growth resistance curves, revealing no significant change of crack closure mechanisms between T6 and HIP condition. But the load stress ratio shows a significant impact on the contributions of crack closure mechanisms to the long crack threshold. The experimental data indicates that common mean stress sensitivity models can lead to non-conservative fatigue design.Finally, a mean stress sensitivity model for fatigue strength assessment is set up, merging existing approaches of the linear elastic fracture mechanics and nominal stress fatigue assessment methods to a three dimensional fatigue assessment map. The introduced approach takes both the present critical defect size for a given probability of occurrence, as well as the decreasing fatigue strength and diminishing contributions of crack closure effects with increasing load stress ratio R into account. The model is validated by for HIP and T6 post-treated material under alternating and tumescent test conditions, revealing an average deviation of only three percent to the experimental data.