The framework of geweke (1982. Journal of the american statistical association 77, 304–324.) and hosoya (1991. Probability theory and related fields 88, 429–444.) is adopted to construct a simple test for causality in the frequency domain. This test can also be applied to cointegrated systems. To study the large sample properties of the test, we analyze the power against a sequence of local alternatives. The finite sample properties are investigated by means of monte carlo simulations. Our methodology is applied to investigate the predictive content of the yield spread for future output growth. Using quarterly us data we observe reasonable leading indicator properties at frequencies around one year and typical business cycle frequencies.