The aim of the study was to evaluate reproducibility of the dGEMRIC (delayed Gadolinium Enhanced Magnetic Resonance Imaging of Cartilage) method in knee cartilage between I and II measurements compared to II and III with ten asymptomatic adults. The dGEMRIC experiment was repeated for tibiofemoral and patellar cartilages three times with an average interval of five days between scans. The measurements were performed by four technicians, whose turns of making the measurements were coincidental during the study. At first, the participants went through clinical MRI-series. Then a Gadolinium-oriented contrast agent was intravenously injected. After 90 minutes penetration time, the T1 weighted GEMRIC imaging was made according the clinical series` landmarks. Particular ROI (region of interest) segments and their reproducibility were separately analysed and computed from femoral, tibial and patellar cartilages. Also the bulk values for entire cartilage were computed. The reproducibility between the measurements were analysed on the group level by Coefficient of Variation (CVrms) and Intraclass Correlation Coefficient (ICC). The reproducibility was moderate or good between the I and II measurements (CVrms 4.1-10.2 %, ICC 0.47 to 0.98) and between the II and III measurements (CVrms 3.3–13.6 %, ICC 0.47 to 0.99) in all ROI segments. The reproducibility was slightly lower in the bulk values both between the I and II measurements (CVrms 9.9– 16.6 %, ICC 0.86 to 0.95) and the II and III measurements (CVrms 9.8 – 17.8 %, ICC 0.88 to 0.98). According to the results, the reproducibility of dGEMRIC method can be mainly considered good in the segments of femoral, tibial and patellar cartilages both in deep and in superficial layer regardless of the variation of technicians.