For a significant range of offshore wave conditions and foreshore slopes, run-up observations are compared to semiempirical formulations and predictions of an existing numerical model based on the depth-averaged one-dimensional nonlinear shallow water equations with bore-like breaking wave dissipation and quadratic bottom friction. The numerical model is initialized with time series of sea surface elevation and cross-shore velocity observed in 80 cm mean water depth (approximately 50 m offshore of the mean shoreline) on a gently sloping beach and in 175 cm water depth (100 m offshore of the shoreline) on a steep concave beach. Run-up was measured with a stack of resistance wires at elevations 5, 10, 15, 20, and 25 cm above and parallel to the beach face. At sea swell frequencies (nominally 0.05<f≤0.18 Hz), run-up energy is limited by surf zone dissipation of shoreward propagating waves so that increasing the offshore wave height above a threshold value does not substantially increase the predicted or observed sea swell run-up excursions (e.g., run-up is ''saturated''). Existing semiempirical saturation formulations are most consistent with the observations and numerical model predictions of run-up excursions nearest the bed. In contrast, at infragravity frequencies (0.004<f≤0.05 Hz) where surf zone dissipation is relatively weak and reflection from the beach face is strong (e.g., saturation formulas are not applicable), the run-up excursions increase approximately linearly with increasing offshore wave height. The numerical model also accurately predicts that the tongue-like shape of the run-up results in sensitivity of run-up measurements to wire elevation. For instance, run-up excursions and mean vertical superelevation (above the offshore still water level) increase with decreasing wire elevation, and continuous thinning of the run-up tongue during the wave uprush can result in large phase differences between run-up excursions measured at different wire elevations. Numerical model simulations suggest that run-up measured more than a few centimeters above the bed cannot be used to infer even the sign of the fluid velocities in the run-up tongue. ¿ American Geophysical Union 1996