Technological heterogeneity is an important issue in studies of agricultural production. We assume that this is much more serious in transition countries, since the agricultural sector in these countries is characterized by the presence of even more different technologies and structures. Previous studies address the issue of production heterogeneity in developed countries; however, there is a clear lack in the literature concerning the effect of different technologies in transition countries. There are two common approaches to estimate different technologies: the most common one is to split the sample into groups based on some a priori information; an alternative method is the latent class models. In the present paper both approaches are used to identify different technologies and to estimate technical efficiency. It seems to be that the LCM model identified better technological differences and separated better the effect of heterogeneity and technical efficiency.