This paper proposes a procedure for solving a security constrained optimal power dispatch (SCOPD) problem under normal and emergency conditions using ant colony optimization (ACO) algorithm. The objective function is to minimize the non-linear generation cost function by optimizing the control variables of the generators real power under equality and inequality constraints. The ACO algorithm is applied to 5- bus system, the IEEE standard 14-bus and 30-bus systems. In addition, an application of the proposed algorithm to a real power system at the west Delta network (WDN) as a part of the Unified Egyptian Network (UEN) considering the valve-points effects has been demonstrated. The results obtained are compared with those obtained using a conventional linear programming (LP), the fuzzy linear programming (FLP) technique and genetic algorithm (GA). Simulation results show that the proposed ACO algorithm for the SCOPD is more accurate and efficient, especially with increasing the system size.