The aim of this work was the application of time series analysis theory and SAS software to select and realize forecasting methodic of Lithuanian monetary, inflation, interest rates and foreign trade indicators and estimate errors of forecast.
Trends and seasonal indices of main economic indicators where estimated. The models of trends were mainly linear but for CPI (consumer price index) and interest rates were used non-linear models. For estimation of seasonal index simple add-up method and procedure Xll were applied. The random fluctuations of indicators were circumscribed using ARIMA models. Applied one-yearahead forecast confirmed the expediency of auto regressive models for Lithuanian economics research.