We examine the effects of individual, team, and institutional capabilities on the governance of technology contracts. Star scientific teams may work on higher quality projects which may be of high or low risk, depending on the maturity of the technology. Arguments that assumed that both capabilities and risk are codetermined, and seldom diverge in their effects on incentive preferences, may be tenuous in these cases. We test our predictions using a two-stage model in a sample of 1,474 inventions that were licensed through performance or upfront contracts. We find that when individuals and teams have strong capabilities, they prefer performance contracts, even if the underlying technology was more mature (or of lower risk). These results explain conditions when inventors facing the same technology risk elect to organize their transactions differently. The implications for theories of risk, capabilities, and innovation are discussed.