The low power of available econometric tests is an important problem in applied research on unit roots and related issues. Based on the principle of methodological triangulation, the problem should be analyzed from different points of view in order to increase the validity of the results. Following this approach a strategy to test the order of integration in time series is presented using a sequence of eleven consolidated tests. In this way it is possible to determine the persistence of shocks, to specify the best strategy for trend-cycle decomposition and to obtain additional information useful for public policies. As an application of the methodology, the integration properties in the main 14 Argentine macroeconomic variables are studied. A classification of them in four homogenous groups according to their order of integration is obtained.