This paper presents a model aimed at finding an efficient allocation of infrastructure investments in a region. The problem's complexity is due, not only to its combinatorial nature, but also due to the intrinsic multidimensional spatio-temporal relationships of its variables. Furthermore, there is no explicit solution for such NP-complete combinatorial optimisation problem; thus a heuristic optimisation technique such as Simulated Annealing is used to search for ”good" solutions in a finite but huge solution space. In this paper, the approach applied in the “Xuzhou Integrated Settlement and Transportation Planning Project", carried out in the People's Republic of China as a joint venture between the Jiangsu Development Planning Commission (JDPC) and the Institute of Regional Development Planning of the University of Stuttgart (IREUS), is to be presented. This study considered projects in 18 realms of infrastructure, in 115 locations of an administrative unit with about 9 million inhabitants. The results of the study suggest a significant gain in allocation efficiency due to the applied method of optimisation. Keywords: Infrastructure location, combinatorial optimisation, Simulated Annealing.