Knowledge of whether, and over what range of output, there are economies or diseconomies of scale in providing local communal services is an important question from theoretical, practical and regional political point of view also. The theoretical side of the question is connected to the primordial research concerning to the optimal city size. If optimal city size actually can be established, then a valid policy argument can be made for fostering its approximation. However, theoretical considerations are based sometimes those types of assumptions, which are often not valid in reality. Therefore empirical investigations are essential in this research area. After a short theoretical review, our paper firstly gives a general outline about the previous controversial empirical evidences on economies of scale in the community size and providing local communal services. Various conceptual and comparability problems in this issue are also examined, such as the differences between technical and monetary measures, the measurement problems of costs and outputs, the effect of population density, spatial extension of the community, the integrated service provision and the quality differences of services, the difference between service plant level and community level investigations. In the second part we present our empirical findings concerning to economies of scale in local communal services based on a large and detailed ten years long data base which consist of more than 200 Hungarian towns above 5 thousand inhabitants. The analysis has two levels, the community level and service level. On community level the general per capita costs are compared for the whole community, on service level the individual services are the observational units. During the examination several methodological questions have occurred. For example, there are several solutions for the organizational structure of providing local services, from the big holding to the smaller individual companies. Considering this and some other issues, the main results suggest that on community level there is a moderated economies of scale until 20 thousand inhabitants, but above this level there is not connection between the settlement size and the average cost of services. This result is mainly consistent with the previous findings: there are economies of scale under a threshold, but after reaching this, unit cost reduction is not feasible. This threshold is different in the different types of services.
