Restoration of degraded wet meadows found on upland valley floors has been proposed to achieve a range of ecological benefits, including augmenting late‐season streamflow. There are, however, few field and modelling studies documenting hydrologic changes following restoration that can be used to validate this expectation, and published changes in groundwater levels and streamflow following restoration are inconclusive. Here, we assess the streamflow benefit that can be obtained by wet‐meadow restoration using a physically based quantitative analysis. This framework employs a 1‐dimensional linearized Boussinesq equation with a superimposed solution for changes in storage due to groundwater upwelling and evapotranspiration, calculated explicitly using the White method. The model and assumptions gave rise to predictions in good agreement with field data from the Middle Fork John Day watershed in Oregon, USA. While raising channel beds can increase total water storage via increases in water table elevation in upland valley bottoms, the contributions of both lateral and longitudinal drainage from restored floodplains to late‐summer streamflow were found to be undetectably small, while losses in streamflow due to greater transpiration, lower hydraulic gradients, and less laterally drainable pore volume were likely to be substantial. Although late‐summer streamflow increases should not be expected as a direct result of wet‐meadow restoration, these approaches offer benefits for improving the quality and health of riparian and meadow vegetation that would warrant considering such measures, even at the cost of increased water demand and reduced streamflow.