This paper attempts to evaluate the information content of money for the forecast of inflation, output, investment and consumption in the euro zone. It considers M1 and M3; a number of modifications to these aggregates is also proposed to enhance their forecast performance. The evaluation employs Granger-causality tests, stability tests and historical out-of-sample forecasts. On balance the information content of money appears to be rather limited. An improvement of the forecast is confined to the real variables and to the second half of the nineties. For the first half of the nineties the forecast performance of money is generally poor.