In the area of time series analysis and prediction there are comparatively many methods available, out of which the extrapolation method has been practically applied most often. Currently, combined models have been serving more and more in experimentation. This study is aimed at construction of adequate models of the indicators observed development tendencies, assessment of selected individual models and subsequent aggregation of these into combined models, including a comparative analysis of both types of models. To find suitable candidates for predicting within the time series analysed, SAS system has been used. Outcomes of the empirical study have shown promising results in the use of combined models for time series processing. The techniques presented in the paper are illustrated with examples of short-term time series on monthly and quarterly basis, in the field of mobile telecommunication services, their consumption and use in various regions of the Czech Republic. A description of the current state of use of selected services in separate CR regions, including urban and rural areas, is a natural part of the paper. ICT including mobile phones and use of mobile services, especially in rural areas, is still widely discussed topic. The research was prepared with the support of the Internal Grant Agency of the Faculty of Economics and Management of the Czech University of Life Sciences, within the project „Methodological Approaches to Identify Economically Weak Regions in the Border Areas of the Czech Republic“(No. 20151035).