The consistency of the heading expressions used in each screen in a plurality of evaluated screens is exhaustively and reliably evaluated. The GUI evaluation system comprises: GUI information storage means for storing GUI information that concerns heading included in an evaluation target screen and includes information indicative of heading expression which is the expression used for the heading; heading group specification means for grouping headings included in each evaluation target screen by expression used for the headings in accordance with the GUI information stored in the GUI information storage means; and heading expression evaluation means for evaluating a consistency of heading expressions between a plurality of evaluation target screens by comparing heading groups that are grouped by the heading group specification means and included in all possible combinations of two of the plurality of evaluation target screens.