This dissertation investigates multi-period salesforce incentive contracting. The first chapter is an overview of the problems as well as the main findings. The second chapter continues with a review of the related literatures. The third and fourth chapters address a central question in salesforce contracting: how frequently should a firm compensate its sales agents over a long-term horizon? Agents can game the long-term contract by varying their effort levels dynamically over time, as discussed in Chapter 3, or by altering between a “bold" action and a “safe" action dynamically over time, as discussed in Chapter 4. Chapter 3 studies multi-period salesforce incentive provisions when agents are able to vary their demand-enhancing effort levels dynamically. I establish a stylized agency-theory model to analyze this central question. I consider salespeople's dynamic responses in exerting effort (often known as “gaming"). I find that long time horizon contracts weakly dominate short time horizon contracts, even though they enable gaming by the agent, because they allow compensation to be contingent on more extreme outcomes; this not only motivates the salesperson more, but also leads to lower expected payment to the salesperson. A counterintuitive observation that my analysis provides is that under the optimal long time horizon contract, the firm may find it optimal to induce the agent to not exert high effort in every period. This provides a rationale for effort exertion patterns that are often interpreted as suboptimal for the firm (e.g., exerting effort only in early periods, often called “giving up"; exerting effort only in later periods, often called “postponing effort"). I also discuss the implication of sales pull-in and push-out, and dependence of periods (through limited inventory) upon the structure of the optimal contracting. Chapter 4 examines multi-period salesforce incentive contracting, where sales agents can dynamically choose between a bold action with higher sales potential but also higher variance, and a safe action with limited sales potential but lower variance. I find that the contract format is determined by how much the firm wants later actions to depend on earlier outcomes. Making later actions independent of earlier demand outcomes reduces agents' gaming, but it also reduces an agent's incentive to take bold actions. When the two periods are independent, an extreme two-period contract with a hard-to-achieve quota, or a polarized two-period contract allowing agents to make up sales, can strictly dominate a period-by-period contract, because they induce more bold actions in earlier periods by making later actions dependent on earlier outcomes. However, when the two periods are dependent through a limited inventory to be sold across two periods, the period-by-period contract can strictly dominate the two-period contract, by allowing the principal more flexibility in adjusting the contract.