Abstract Purpose K-edge imaging based on the photon counting detectors (PCDs) is an effective enhanced imaging method because the PCDs are conducive to the K-edge imaging due to the adjustable energy thresholds. The energy bins significantly affect the image quality of the K-edge imaging, but the conventional energy bins used for K-edge imaging are continuous which weaken the K-edge signal and decline the image quality. Hence, how to get a better K-edge signal by the optimized energy bins is the key point for the K-edge imaging based on the PCDs. Method This paper experimentally studied the influence of the energy bins used for the K-edge imaging based on the PCDs. The conventional energy bins were determined by the theoretical-attenuation method (TAM), and the optimized energy bins were determined by the threshold-scan method (TSM). For the phantom and mice imaging, we performed both the K-edge subtraction algorithm and the K-edge decomposition algorithm on the projections obtained by the energy bins which were determined by the TAM and TSM. The image quality was compared using the CNR of the objective area. Results The experimental results showed that the energy bins identified by the TSM had a better performance than the TAM in both imaging methods. The TSM improved the CNR by ~ 39% than the TAM in the phantom results and could better highlight the areas where the contrast agents are enriched (such as the kidney). Conclusions The optimized energy bins can better highlight the K-edge signal than the conventional energy bins which can improve the image quality and have the potential to reduce the amount of the contrast agents