Standard RBC models fail to account for the business cycle stylized facts when growth results from a random walk in technology. Does this mean that we are to dismiss the evidence in favor of a unit root in output when we have a prior in favor of RBC models? To answer this question, we explore the usefulness of introducing an endogenous source of growth in a standard RBC model. Using a formal measure of Þt, we show that doing so permits both to reproduce some key business cycle facts and to obtain a unit root in output.