A hydrodynamic model incorporating a self-consistent treatment of ocean self-attraction and loading (SAL), and a physically based parameterization of internal tide (IT) drag, is used to assess how accurately barotropic tides can be modeled without benefit of data, and to explore tidal energetics in the last glacial maximum (LGM). M2 solutions computed at high resolution with present day bathymetry agree with estimates of elevations from satellite altimetry within 5 cm RMS in the open ocean. This accuracy, and agreement with atlimetric estimates of energy dissipation, are achieved only when SAL and IT drag are included in the model. Solutions are sensitive to perturbations to bathymetry, and inaccuracies in available global databases probably account for much of the remaining error in modeled elevations. The ≈100 m drop in sea level during the LGM results in significant changes in modeled M2 tides, with some amplitudes in the North Atlantic increasing by factors of 2 or more. Dissipation is also significantly changed by the drop in sea level. If IT drag estimated for the modern ocean is assumed, dissipation increases by about 50% globally, and almost triples in the deep ocean. However, IT drag depends on ocean stratification, which is poorly known for the LGM. Tests with modified IT drag suggest that the tendency to a global increase in dissipation is a robust result, but details are sensitive to stratification. Significant uncertainties about paleotides thus remain even in this comparatively simple case where bathymetry is well constrained.