This study assesses the usefulness of money for policy in the Philippines. The basic idea behind the use of monetary aggregates for policy is that observed fluctuations in money anticipate movements in the ultimate objective of monetary policy, such as inflation control. The paper examines the stability of key empirical relationships, including the behavior of velocity and the presence of cointegrating relationships among money and variables of interest to policymakers. In general, results indicate that the stability of velocity and the presence of cointegrating relationships lend some limited support to the potential usefulness of money for policy. The ability of money to predict inflation is examined using Granger causality tests and an unrestricted vector autoregression (VAR) that examines the relative contribution of innovations in money to the variance of the forecast errors in inflation. In general, money's ability to predict inflation is less clear-cut and seems to be dependent on the ordering and lag lengths of the variables used in the VAR and the definition of money used.