This article sheds light on how industry fluctuations affect firms' propensity to innovate. We test two seemingly conflicting arguments that suggest how firms are more or less inclined to engage in innovation activities during industry fluctuations. By studying a panel of 622 Italian manufacturing firms during the period 1995-2003, we show how differentiating between product and process innovation may help reconcile the theory of opportunity cost of innovation with the cash-flow effect argument. We find that industry downturns are related to product and process innovation in different ways: firms tend to invest in product innovation rather than process innovation in downturns. The findings have implications for both theory (showing when the opportunity cost of innovation dominates) and research design (showing the importance of both the input and output measures in innovation studies and how they might influence the results).