In this paper, we use a wavelet filtering based approach to study the econometric relationship between exports, imports, and economic growth for Tunisia, using quarterly data for the period 1961:1-2007:4. GDP is used as a proxy for economic growth. We explore the interactions between these primary macroeconomic inputs in a co-integrating framework. We also study the direction of causality between the three variables, based on the more robust Toda-Yamamoto modified Wald (MWALD) test. The much-studied relationship between these three primary indicators of the economy is explored with the help of the wavelet multi-resolution filtering technique. Instead of an analysis at the original series level, as is usually done, we first decompose the variables using wavelet decomposition technique at various scales of resolution and obtain relationship among components of the decomposed series matched to its scale. The analysis reveals interesting aspects of the interrelationship among the three fundamental macroeconomic variables.