Purpose: Hybrid Photon Counting (HPC) detectors profoundly improved x-ray diffraction experiments at third generation synchrotron facilities. Enabling the simultaneous measurement of x-ray intensities in multiple energy bins, they also have many potential applications in the field of medical imaging. A prerequisite for this is a clean spectral response. To quantify how efficiently HPC detectors are able to assign photons to the correct energy bin, a quantity called Spectral Efficiency (SE) is introduced. This figure of merit measures the number of x-rays with correctly assigned energy normalized to the number of incoming photons. Methods: A prototype HPC detector has been used to perform precision measurements of x-ray spectra at the BESSY synchrotron. The detector consists of a novel ASIC with pixels of 75 9 75 lm 2 size and a 750 lm thick CdTe sensor. The experimental data are complemented by the results of a Monte-Carlo (MC) simulation, which not only includes the physical detection process but also pulse pile-up at high photon fluxes. The spectra and the measured photon flux are used to infer the Spectral Efficiency. Results: In the energy range from 10 to 60 keV, both the Quantum Efficiency and the Spectral Efficiency were precisely measured and simulated. Good agreement between simulation and experiment has been achieved. For the small pixels of the prototype detector, a SE between 15% and 77% has been determined. The MC simulation is used to predict the SE for various pixel sizes at different photon fluxes. For a typical flux of 5•10 7 photons/mm 2 /s used in human Computed Tomography (CT), the highest SE is achieved for pixel sizes in the range between 150 9 150 lm 2 and 300 9 300 lm 2 . Conclusions: The Spectral Efficiency turns out to be a useful figure of merit to quantify the spectral performance of HPC detectors. It allows a quantitative comparison of detectors with different sensor and ASIC configurations over a broad range of x-ray energies and fluxes. The maximization of the SE is a prerequisite for a successful usage of HPC detectors in the field of medical imaging.