This article presents a computationally efficient technique for the macroscopic (equation-based) network loading of microscopic (simulation-based) travel demand. It combines the expressive power of microsimulations as sampling tools for highly heterogeneous demand segments with mathematically well-understood macroscopic network loading procedures. The presented experimental results demonstrate the efficiency and precision of the approach even for large and complex scenarios. This work is intended to contribute to the understanding and application of microsimulation techniques in the context of mathematically motivated dynamic traffic assignment procedures.