GRADD is a group decision support systems targeted at asynchronous and distributed meetings. The distinctive features of GRADD are the combination of a rationalemodel for discussion support, voting techniques for deliberation activities, and automatic norm control for process structure definition and control. The goal of this paper is to present the results of an empirical evaluation of GRADD functionality, with regard to these distinctive features. The experiment, though empirical, enabled to detect the strong and weak points of the proposed functionality and/or of the current GRADD prototype, and to establish future extensions and improvements.