Optimum design of real world steel space frames under design code provisions is a complicated optimization problem due to the presence of large numbers of highly nonlinear constraints and discrete design variables. The use of gradient based optimization techniques in finding the optimum solution of such large design problems is cumbersome due to the selection of initial design points and convergence difficulties while meta heuristic algorithms do not suffer such problems. Artificial bee colony (ABC) algorithm is one of the recent additions to the swarm intelligence based meta-heuristic search techniques that mimic natural foraging behavior of honey bees. In this study optimum design problem of steel space frames is formulated according to the provisions of LRFD-AISC and its solution is obtained by using enhanced artificial bee colony algorithm. The performance of artificial bee colony algorithm is improved by adding Levy flight distribution in the search of scout bees. Real world steel space frames are designed with the new algorithm developed in this study to demonstrate its robustness and efficiency. (C) 2015 Elsevier Ltd. All rights reserved.