This study presents a design-driven heuristic approach named guided stochastic search (GSS) technique for discrete sizing optimization of steel trusses. The method works on the basis of guiding the optimization process using the well-known principle of virtual work as well as the information collected during the structural analysis and design stages. The performance of the proposed technique is investigated through a benchmark truss instance as well as four real-size trusses sized for minimum weight according to AISC-LRFD specifications. A comparison of the numerical results obtained using the GSS with those of other available algorithms indicates that the proposed technique is capable of locating promising solutions using lesser computational effort. (C) 2014 Elsevier Ltd. All rights reserved.