Abstract: Traditional single-energy cone-beam computed tomography (CT) imaging could provide only structural information and retains poor quantitative calculation performance. If the scanned object contains high-density materials (e.g., metal implants), significant metal artifacts would appear in the single-energy cone-beam CT reconstructed images, affecting clinical diagnosis. In this study, we used two dual-energy cone-beam CT systems to scan a human forearm and a dental phantom with metal implants, achieving spectral cone-beam CT imaging through single-source sequential scanning and kilovoltage switching. By comparing the single-energy cone-beam CT reconstructed images with the spectral cone-beam CT reconstructed images, we verified that spectral cone-beam CT exhibits good quantitative calculation performance and metal artifact reduction effects. In regions of the human forearm model with different hydroxyapatite (HA) concentrations, spectral CT imaging bone density quantitative calculation error could remain within 1 %. In virtual monoenergetic images at high energy levels, the metal artifact-induced non-uniformity could be reduced from nearly 600 Hounsfield Units (HU) to within 20 HU. This study demonstrates the excellent performance of spectral cone-beam CT in quantitative calculation and metal artifact reduction, providing a reference for subsequent clinical applications.