In the development of advanced data acquisition, testing or telecommunication equipment, one often relies on modular off-the-shelf processor boards and I/O-boards that are being composed to a single distributed system. To support such architectures, the microTCA (micro Telecommunication Computing Architecture) offers standardized racks and standardized back panel communication protocols connecting the different boards. During the design of such distributed systems, the performance characteristics of the end-to-end communication between different boards in the system are often not sufficiently known. In this paper, we present a setup allowing for benchmarking the performance of the full PCIe-communication path between two microTCA FPGA-boards. The experimental setup is discussed, and it is shown that benchmarking figures for throughput, delay and delay variation of the PCIe end-to-end path can be retrieved.