Photon counting applications are typically limited by detector deadtime to operate at count rates of a few megahertz, at best, and often at significantly lower levels. This limitation is becoming more critical with the advance of photon counting applications such as photon-based quantum information. We present a first experimental proof of principle, and review the theoretical foundation of a multiplexed detection scheme that allows photons to be counted at higher rates than is possible with individual detectors or simple detector trees. In addition to this deadtime improvement, we discuss the impact of this scheme on other relevant characteristics such as afterpulsing and dark count rates.