This paper explores the underlying causes of volatility in R&D performance over time at the firm level. R&D performance volatility has not been deeply examined in the innovation literature despite the fact that it plays a critical role in industries such as pharmaceuticals or the movie industry, where firms often undergo ?hot? and ?cold? streaks in R&D output. In this paper, we use a simulation model to explore such phenomenon, building on insights from behavioral theories of the firm: we argue that the swings in performance, while rooted in uncertainty, are exacerbated by the behavioral influences in how decision makers deal with risk and uncertainty in R&D. In particular, we propose and formally argue that a firm can affect both R&D output and R&D volatility by modifying its behavior toward risk or by changing the policy it uses to manage its projects portfolio. We simulate a multi-stage R&D process characterized by uncertainty in lead times and projects success. We explore the effect of scale, resource allocation strategies, risk preferences and behavior towards available information on R&D output and volatility. Our results highlight an important tradeoff firms face when managing their R&D portfolio, i.e., the tradeoff between output and volatility. We find that given a fixed budget, priority rules that assure higher R&D outputs come with a ?burden?: higher volatility. Analogously, we show that the same tradeoff between output and volatility exists when the changing policy regards firms? risk preferences instead of prioritization rules.