Abstract: This dissertation consists of two essays on contract theory. The first essay considers the role of costly and imperfect courts of law for the enforcement of contracts in the hold-up problem. The second considers optimal contracting in a principal-agent model in which the agent can acquire hard information about his costs before agreeing to a contract. \"Hold-Up under Costly Litigation and Imperfect Courts of Law\" . The classic hold-up literature establishes that without a contract, two parties in a relationship will underinvest in cost-reducing or quality-enhancing activities that are relationship specific. Recent studies have asked whether this hold-up inefficiency is alleviated by contracting, under two stipulations: (1) the investments cannot be contracted upon because, e.g., they are unverifiable, and (2) the parties are unable to commit themselves to avoid renegotiation. Two main results have been obtained. First, a simple contract specifying a price and quantity of the final good to be traded will, fairly generally, induce efficient investments if the investments are \'selfish\' in nature, i.e., each party\'s investment directly affects only his own profit (Edlin and Reichelstein, 1996). Second, and in contrast, no contract however complicated is of any value in reducing the inefficiency if the investments are \'cooperative\' in nature, i.e., each party\'s investment affects directly only the other party\'s profit (Che and Hausch, 1999). This essay shows that courts of law may play a more important role in real contract disputes than has been realized. The key observation is that the presence of a court can make it valuable to specify putative investment levels in a contract--even if the court remains ignorant of the parties\' actual investment levels. This is because the putative investment levels affect the expected damages the court awards if it decides that breach occurred. The probability of the court deciding breach occurred is independent of the actual investment levels--they remain entirely unverifiable. This probability depends at most on the parties\' court expenditures. These expenditures make litigation costly for the parties, and so in equilibrium they settle before going to court. The presence of even such an imperfect court has a significant impact on whether contracts alleviate hold-up inefficiencies. In the case of one-sided cooperative investment, I show that the first-best outcome can sometimes be achieved by the adoption of a simple non-contingent contract, contrary to the negative result of Che and Hausch (1999). Our result extends to the case of hybrid investment, provided the investment is mainly cooperative. \"Pre-Contractual Information Acquisition\" . I consider a principal-agent model in which the agent may acquire costly information about his effort costs before accepting a contract. The model departs from the literature in two ways: (1) the information is \'hard\' in the sense that it can be credibly communicated, and (2) the parties are unable to commit themselves not to renegotiate. When the cost of acquiring information is low, the optimal contract induces information acquisition. In this case the contract is renegotiated and leaves the agent no rent. When the information cost is higher, the optimal contract induces the agent to not acquire information. In this case, if the cost is not too high, quantities are inefficient and the agent may receive rent. If the cost is yet higher, the contract again yields efficient quantities and leaves the agent no rent. These results hold also if parties can commit themselves not to renegotiate.