This paper presents the role of performance parameters for an aircraft gas turbine engine in influencing the configuration of systems/sub-systems and related costs at the preliminary design phase. An expert decision support system is provided that employs a hybrid decision tree inference mechanism to provide cost estimates for configurations of a system/sub-system based on the specifications of the performance parameters by the user. The inference mechanism provides a way of dealing with certain and uncertain knowledge extracted from experts for decision making. This assists the designer in the preliminary phase of design in evaluating decisions on the specification of performance parameters by considering their cost influence. The designer can optimize the system/sub-system configuration and performance parameters based on cost. A case study demonstrating the research applied to an exemplar turbojet engine compressor sub-system is presented.