This talk gives a gentle introduction to cointegration models. A vector of time series is cointegrated if each is nonstationary with a unit root (i.e., they are first difference stationary), but a linear combination of them produces a stationary time series. More mathematically, a vector of time series yt is cointegrated if each element in is I(1), but there exists a vector a such that a’yt is I(0). We begin by exploring the spurious regression phenomenon, distinguishing these relationships from a cointegration relationship. We then introduce the cointegration model and discuss hypothesis testing for the null hypothesis of no cointegration as well as hypotheses regarding the cointegration vector.