Michael T. Belongia and James A. Chalfant review the arguments about why M l velocity may have changed, then discuss more general issues about the construction of a measure of the money stock and the properties one would expect it to exhibit. Noting that the composition and weighting of components in a monetary aggregate each may contribute to measurement error and unexplained shifts in behavioral relationships, the authors then review other recent efforts to construct new monetary aggregates from first principles of economic and statistical theory. They compare the empirical performance of the alternatives against various final use criteria and find that, although some measures offer improvements over the current official aggregates, none can be judged at this time to be the definitive measure of the money stock.