A mathematical terrain model, from which it is possible to predict the number of valleys visible from the highest elevation within an area, has been applied to the New England physiographic province and to an area in the Colorado Plateau. Correlation between the predicted and actual number of visible valleys for two areas is good.