We study how people with present bias make choices when they face multiple projects. Each project consists of a starting and a finishing stage, both requiring costly effort to complete but yield rewards only after project completion. Methodology/results: We analytically derive people's perception-perfect strategies for project scheduling. Naive people (naifs), who are unaware of their self-control problems, may start a project, but never finish it. They may multitask; that is, start a new project before finishing an old project. They may also start a project with a lower net present value (NPV) before they start one with a higher NPV. These behaviors are suboptimal from a long-run perspective. Sophisticated people (so-phisticates), who are aware of their self-control problems, do not start a project but not finish it. However, like naifs, they may multitask and not prioritize projects on the basis of NPV. Multitasking is more likely when the projects all have low start-up costs and the projects with the lowest start-up costs have high costs to finish. People are more likely to start low NPV projects first if high NPV projects have high start-up costs. What happens if people are allowed to choose the cost structure endogenously? In this case, if people multitask when the cost structure is exogenously given, then when the cost structure can be chosen by them, naifs delay the costs to the finishing stages as much as possible, resulting in either multitasking or procrastination. If, in addition, the project with a higher NPV has a lower total cost, then naifs prioritize projects by NPV. Sophisticates, however, may or may not delay the costs to finishing stages when given a choice because delaying the costs may lead to multitasking, which decreases the * Corresponding author 1 utility and they, unlike naifs, are fully aware of. Therefore, under endogenous cost structure, sophisticates may or may not multitask or prioritize projects by NPV. Managerial implications: We demonstrate that enhancing individuals' awareness of their present bias (turning naifs into sophisticates) makes them less likely to behave sub-optimally. Increasing load (adding projects to project portfolios) can also, though not always, alleviate procrastination and reduce the possibility of multitasking and not prioritizing projects based on NPV. When project portfolios are predetermined, increasing awareness of present bias always improves people's long-run utility. However, when project portfolios are selected endogenously, sophistication does not always pay because naifs may either complete more projects or complete a more profitable project than sophisticates.