If monetary policy is to be effective in controlling the macroeconomy, accurate measurement of the money supply is essential. The conventional way of measuring the level of the money supply is to simply sum the constituent liquid liabilities of banks. However, a more sophisticated, weighted monetary index has been proposed to take account of the varying degrees of liquidity of the short-term instruments included in money. Inferences about the effects of money on economic activity may depend importantly on the choice of monetary index because simple sum aggregates cannot internalize pure substitution effects. This hypothesis is investigated in the current paper. A Divisia index measure of money is constructed for the USA, UK and Italian economies and its inflation forecasting potential is compared with that of its simple sum counterpart in each of the three countries. The powerful Artificial Intelligence technique of neural networks is used to allow a completely flexible mapping of the variables and a greater variety of functional form than is currently achievable using conventional econometric techniques. The application of neural network methodology to examine the money-inflation link is highly experimental in nature and, hence, the overriding feature of this research is one of simplicity. Superior inflation forecasting models are achieved when a Divisia M2 measure of money is used in the majority of cases. This support for Divisia is entirely consistent with findings based on standard econometric techniques reported from the respective central and Federal Reserve banks of each country. Divisia monetary aggregates appear to offer advantages over their simple sum counterparts as macroeconomic indicators. Further, the combination of Divisia measures of money with the artificial neural network offers a promising starting point for improved models of inflation.