Parallel robots offer an advantageous solution for large and complex structural components, and their kinematic performance serves as an important indicator of operational capability. However, the evaluation of parallel robots involves a wide variety of kinematic performance indices, and the mapping relationships between these indices and dimensional parameters differ from one another, making dimensional design that simultaneously accounts for multiple kinematic performances highly challenging. Aiming to comprehensively characterize the kinematic performance of parallel robots by integrating instantaneous and continuous performance measures, this study established a unified performance characterization model, thereby reformulating the complex dimensional parameter - kinematic performance matching problem as a parameter optimization problem. Firstly, with the expected threshold values of various kinematic performance indices taken as the boundary conditions, a comprehensive motion capability index for parallel robots was formulated subject to footprint constraints, and a dimensional optimization model was established by taking this index as the design objective. Then, based on the proposed model, a dimensional synthesis strategy for parallel robots that considered both instantaneous and continuous kinematic performances was developed, and an improved Monte Carlo method was employed to achieve rapid optimization of dimensional parameters. Finally, the 2PRU-(2PRU) R redundantly actuated parallel robot and the 3P(2UU) exactly actuated parallel robot were taken as two design cases to demonstrate the dimensional design procedure considering multiple kinematic performances, and a comprehensive kinematic performance evaluation was conducted to verify the validity of the dimensional design results and the feasibility of the proposed strategy.