In economics and in other fields of life, it is traditional to decompose time series into a variety of components while on the other hand, exponential smoothing is a procedure for continually revising an estimate in the light of more recent experiences. This work set out to compare the forecasting performances of two simple univariate time series analysis, the decomposition and winters’ methods. To achieve this, the methods were applied to stationary and normally distributed data, and stationary time series that   is not normally distributed. In the two data sets considered, the results revealed that the decomposition method outperformed the winters’ seasonal exponential smoothing method. We therefore conclude that both methods are capable forecasting short sample time series, and that the decomposition method forecast better.