Based on historical data from 1980 to 2004 the paper analyzes the main drivers of mail demand and the impact of e-substitution by using time-series techniques. We find strong evidence for e-substitution. In the second part of the paper we provide forecasts for future mail volumes.