<p>A central problem in project management is the optimal use of limited resources. We consider the resource leveling problem, which minimizes the maximum resource load over the life of a project. This problem is motivated by the need to limit expensive resource acquisition. For several variants of this problem, we either provide an efficient optimal algorithm or show that the problem is intractable. We also consider a generalization of the resource leveling problem where work can be subcontracted. The subcontracting cost is defined either by task or by time of subcontracting. Our results show how the solvability of resource leveling problems depends on whether task times are unit or arbitrary, whether resource requirements are unit or binary, whether the project network is a chain, a tree, or an opposing forest, whether subcontracting costs for jobs are unit or binary, and whether the availability of subcontracting at a time period is constrained.</p>