The aim of this study was to investigate the feasibility of delayed cone beam (CBCT) arthrography for clinical diagnostics of knee cartilage lesions. Knee joints with cartilage lesions were imaged using native radiography, MRI, and delayed CBCT arthrography techniques in vivo. The joints were imaged three times with CBCT, just before, immediately after (arthrography) and 45 min after the intra-articular injection of contrast agent. The arthrographic images enabled sensitive detection of the cartilage lesions. Use of arthrographic and delayed images together with their subtraction image enabled also detection of cartilage with inferior integrity. The contrast agent partition in intact cartilage (ICRS grade 0) was lower (p < 0.05) than that of cartilage surrounding the ICRS grade I-IV lesions. Delayed CBCT arthrography provides a novel method for diagnostics of cartilage lesions. Potentially, it can also be used in diagnostics of cartilage degeneration. Due to shorter imaging times, higher resolution, and lower costs of CT over MRI, this technique could provide an alternative for diagnostics of knee pathologies. However, for comprehensive evaluation of the clinical potential of the technique a further clinical study with a large pool of patients having a wide range of cartilage pathologies needs to be conducted. Keywords: cartilage lesions; contrast agent diffusion; computed tomography; delayed CBCT arthrography; osteoarthrosis Cartilage lesions are common and can lead to progressive joint degeneration at high cost for the healthcare system. Early diagnosis and grading of cartilage lesions could improve the early selection of treatment, enable monitoring the effect of new treatments and therefore reduce the risk for post traumatic osteoarthritis (OA).Detection of acute cartilage lesions or early OA is challenging since cartilage is not visible in X-ray imaging. Resolution of magnetic resonance imaging (MRI) is limited, and queuing times are long, but it is still the routine method used in cartilage lesion diagnostics and in clinical decision making for or against arthroscopy. In planar X-ray imaging the Xray attenuation of the cartilage is similar with that of the surrounding soft tissues and therefore cartilage is not visible in the images. For this reason diagnosis is mainly based on decreased joint separation and changes in subchondral bone. On the other hand, MRI is not based on ionizing radiation and can be used to visualize soft tissues enabling detection of cartilage defects. Cone beam computed tomography (CBCT) utilizes a conical X-ray beam, covering a large volume with a single rotation about the patient. Due to differences in imaging geometry and collimation of X-rays the radiation dose of CBCT is much smaller than that of spiral CT.1 MRI is especially useful for diagnostics of meniscal tears and lesions of anterior cruciate ligament (ACL) with accuracies of 81% to 98% and specificities of 69% to 98%, respectively, 2 but in the diagnostics of cartilage lesions its sensitivity and specificity ha...