The co-integration of time series indicates the presence of their common trends. For analytical purpose it is important to transform some time series into gap form. This transformation can be received as a difference between the time series and the common trends. The model of demand for money in Czech Republic created by real M2, real GDP and 1R PRIBOR time series contains two common trends. These trends are estimated from the state space form of unobservable components model. The gap transformations of real M2 and real GDP series can be used for identification of inflation risks. The relatively high correlation was found between gap form of M2 and rate of inflation. The relationship between gap form of GDP and rate of inflation is not so close.