Scott E. Hein and Paul T. W. M. Veugelers examine the accuracy of forecasting velocity growth over the last eight years. The forecasts generated are all ex ante: They utilize only information that was available at the time a forecast was made. The evidence suggests that quarterly velocity growth fluctuates randomly about a fixed level. While it is clear that the magnitude of the quarter-to-quarter fluctuations has increased recently, the available evidence does not indicate that the fixed level has changed. Consequently, the best forecast of future velocity growth is simply the average level of past velocity growth. This simple forecasting approach to velocity growth generally works as well as other, more sophisticated procedures. This article further provides evidence that it is easier to predict average velocity growth over longer time horizons than over shorter ones. The short-run random fluctuations in velocity growth are offsetting in nature so that forecasting accuracy improves when longer-term velocity forecasts are made. This suggests that, other things equal, long-run GNP growth objectives are more likely to be achieved than short-run GNP goals.