Against the background of the recent discussion whether the smoothing behavior of the Fed detected by empirical Taylor rules is indeed a fact or rather a statistically fiction, this paper re-examines the empirical evidence for interest rate smoothing for the case of the European Central Bank (ECB). Based on data representing true ECB behavior, our findings reject the hypothesis of no smoothing but also find a role of serially correlated shocks. The degree of smoothing is estimated in the range of [0.38;0.82], reflecting model uncertainty with respect to the output gap and indicating a rather moderate extent of partial adjustment.