This article deals with some possibilities of the use of time-series analysis for quarterly statistical estimations of macroeconomic aggregates. It is a case of estimations of aggregates of gross domestic product creation and use. Beyond all doubt, their meaning from the point of view of conception and strategy is great. The contemporary standard of these data reporting (either separately or as major measures within national accounts system) in the Czech Republic and, mostly, in further economics under transition starts above all from yearly values of measures. The core of method of their assembling is given by exhaustive surveys. The aim of quarterly statistical estimates of GDP (as well as additional indicators both in vertical and horizontal direction within national accounts system is to provide reliable estimations of quarterly values of corresponding aggregates as soon as possible. This paper deals with statistical macroeconomic model based on indirect methods using short-term surveys. The basic task of indirect methods of quarterly estimations of aggregates of national accounts is, on the one hand, to provide some synthetic information related to just past quarter as soon as possible and, on the other hand, to enable the prediction of corresponding aggregates for 1 - 2 quarters keeping reasonable measure of accuracy and reliability of these predictions.