Recent test programs indicate that theoretical streamlines above a supercavitating hydrofoil may be considerably higher than those actually generated. To check theoretical predictions a three-dimensional traversing probe was constructed and used to map the upper cavity surfaces generated by two different two-dimensional hydrofoils. Experimental contours were found to fall below those predicted by theory at all angles of attack and depths of submergence tested. The rate of change of cavity ordinates with changes in angle of attack or depth of submergence is about the same as that predicted by theory. The contribution of angle of attack to cavity shape is responsible for the greater part of the difference between theoretical and measured cavity shapes. (Author).