Heat source layout optimization is an effective way to enhance heat transfer for electronic cooling. In this paper, the heat source layout optimization in two-dimensional heat conduction is investigated using simulated annealing (SA) method. Mathematical analysis is conducted to transform the heat source layout problem into a combinatorial optimization problem, which can be solved by SA. Three typical cases with various boundary conditions are introduced to validate the effectiveness of SA for heat source layout optimization. The solutions of SA are compared to the ones of random distribution (RD) and the ones of bionic optimization (BO). The results indicate that the maximum temperature of the domain can be remarkably reduced after optimizing the heat source layout using SA compared to RD. Compared to BO, it needs more computational time for SA to obtain the solution. Furthermore, the maximum temperature after optimizing by BO is lower than the ones by SA for the cases with symmetric boundary conditions. While for the case with asymmetric boundary conditions, SA performs better and the maximum temperature lower than BO is obtained. It can be concluded that simulated annealing method is effective to optimize the heat source layout problem in heat conduction. (C) 2016 Elsevier Ltd. All rights reserved.