This paper investigates the robustness of the unit root hypothesis for some of the most representative macroeconomic time series of the Spanish economy covering the period 1954-1998 when structural breaks are allowed to be present. Our analysis shows that, for some variables classified by standard unit root tests as non-stationary time series, the conclusion is reversed when the deterministic trend is specified in such a way that considers the effect of structural breaks.

Key words: unit roots, structural breaks.
JEL classification: C12, C21.