Purpose The objective of this study was to measure resected surfaces of the medial tibial plateau and to provide accurate anatomical parameters for the Chinese population and improve the unicompartmental knee arthroplasty (UKA) component design in the future. Methods This study measured different dimensions of the medial tibial plateau on MRIs of 1000 consecutive healthy Chinese people without knee deformity. The anteroposterior (AP) dimension and mediolateral dimensions at defined points (recorded as WA, WB, WC) of the medial tibial plateau were measured. The aspect ratio (WB/AP, recorded as AR) was also recorded. The measured results were grouped to compare with the dimensions of the Oxford Phase III tibial component. An independent t test was used to compare the differences between the males and females. Results The Oxford Phase III tibial component showed mediolateral overhang or undersize for the whole range of measured anteroposterior dimensions of the resected medial tibial plateau. A total of 71.3% patients' resected surface did not match the Oxford Phase III tibial component well. The study also found a decrease in the aspect ratio (WB/AP) with an increase in the AP dimension in the medial tibial plateau. Conclusions There is a difference between the morphology of the Chinese knee joint and the configuration of the Oxford Phase III tibial components. The results of this study provided accurate anatomical parameters for Chinese and guidelines for designing UKA components suitable for the Chinese population.