The aim of this work is to explain and assess the results of the application of the TRAMO-SEATS seasonal adjustment method on the data of the ISAE manufacturing and consumer surveys. In particular, the study begins by focusing on the description of some of the typical problems of the seasonal adjustment of qualitative series, in relation to the operational choices to be made when applying the procedure (the 'trading day effect', logarithmic transformation of the series, choice of a temporal interval etc.) making the choices explicit for the series of analysis. Subsequently, the characteristics of the seasonal component of the series will be analysed; special attention is given to the identification of the non-stationary seasonality of each series by using a procedure which consists in the extension of a test of the Dickey-Fuller kind to verify the unit roots at the seasonal frequencies. Later, on the basis of the considerations and of the analyses which have been previously made, the models which have been obtained applying Tramo-Seats will then be described highlighting the flexibility of the method in grasping the stochastic characteristics of the seasonality of the series.