In many multi-project organizations, there is a clear need of hierarchical capacity planning. In the early stages of a project, only rough estimates of capacity requirements are available. For this purpose, we introduce the multi-project rough cut capacity planning problem. This problem aims to minimize the use of non-regular capacity, where capacity requirements are only stated in time units (e.g man hours). In addition several constraints must be satisfied, represented by precedence relations, time windows, and minimum durations. We propose two heuristic algorithms and show computational results for problems with different characteristics. Finally, we discuss the potential use of these algorithms for some extensions to the multi-project rough cut capacity planning problem, as defined here.