This paper analyses optimal contracts in a principal-agent model where the agent is intrinsically motivated at the outset and there is an endogenous relationship between the structure of incentive payments and intrinsic motivation (crowding effects). The analysis shows that crowding effects have implications for the optimal contract and that under some conditions the principal can do better without implementing any economic incentives. Furthermore, it is shown that when high-powered incentives diminish instrinsic motivation (crowding-out) the first-best solution in a principal-agent framework is unattainable.