River cross-sections, i.e. riverbed elevation profiles across the river, are required for hydrodynamic modeling in flood risk assessment and flood forecasting but are expensive and time-consuming to survey in the field. Remotely sensed cross-sections can fill the gap in data-scarce areas, with novel satellite missions enabling increased accuracy compared to previous methods. We present a method for estimating river cross-sections from lidar altimetry products and a discharge time series. The exposed cross-section is delineated using high spatial resolution ICESat-2 or airborne lidar point clouds. The submerged section is then inferred using water surface slope from the lidar products together with a discharge input. Assuming the diffusive wave approximation, the energy slope is taken to be equal to the estimated water surface slope. Manning’s equation is subsequently applied to estimate the conveyance at the time of observation. Finally, the full section is defined assuming a roughness parameter and a conceptual submerged shape. The method is implemented in an open-source python package called lidarcross, enabling users to define cross-sections for hydrodynamic models in data-scarce areas. Validation is performed using examples from four surveyed rivers spanning a range of spatial scales. Area and conveyance curves derived from surveyed and lidar-based cross-sections are compared, yielding median WSE differences of approx. 0.5 m. The results demonstrate that the method successfully captures the hydraulic behavior of the channel. Residual errors are primarily attributed to uncertainties in discharge and water surface slope estimates.