This paper examines some techniques that have been applied to estimate the ‘natural’ summer outflow for the Permo-Triassic aquifer of England, the second most important aquifer in the UK for public water supply. A method for naturalizing groundwater models is discussed and the naturalized flows predicted by the groundwater models of three Permo-Triassic Sandstone aquifers are compared against results of an analytical method to determine natural summer outflows for the equivalent aquifer properties. The results from the naturalization of the three groundwater models are broadly comparable and indicate that the simulated naturalized Q85 (i.e. the flow exceeded for 85% of the time) lies in the range of 70–90% of the naturalized long-term average recharge with a mean of 80% long-term average recharge. This provides the basis for a reasonably robust ‘first pass’ method to determine the natural summer outflows of an unconfined Permo-Triassic Sandstone aquifer unit. Some of this variability in the estimated natural summer outflows can be accounted for by the average distance from the ‘upstream’ edge of the aquifer unit to the discharge boundary (i.e. the perpendicular length). Comparison of the results of the naturalized simulations with the predictions of an analytical solution shows that the typical minimum summer baseflow of the analytical solution appears to be most consistent with the Q85 flow condition simulated by the groundwater models. However, it is clear that, at shorter perpendicular lengths (<1.5 km), the simple idealization of the aquifer implicit in the analytical solution is unable to predict natural summer outflows for the Permo-Triassic Sandstone aquifer system. This is likely to be a result of the increased significance of vertical components of flow in smaller aquifer units.