Empirical money demand analysis undertaken at the aggregate level may obscure behavioural differences between the financial, non-financial corporation and household sectors. Looking at the individual and more homogenous sectors may allow more clearly interpretable empirical relationships between money holding, scale variables and opportunity costs to be estimated. Two possible approaches can be taken to address this issue: aggregate and sectoral money holdings are explained either by a common set of determinant variables or by specific determinants, which may differ across sectors. In this analysis, the first approach has been chosen in order to highlight the different elasticities of the long-run money demand with respect to a common set of macroeconomic determinants and thereby to allow comparison of the model for the aggregate M3 with corresponding models for households, non-financial corporations and non-monetary financial intermediaries. This paper presents results for cointegrated VAR systems estimated over a sample of quarterly data from 1991 to 2005. A SUR system is estimated to cross-check the robustness of the findings and to analyse the importance of common shocks across sectors. JEL Classification: E41, C32, E59.