Power divider is a useful device that divides the power of signal into different subpowers at a certain ratio. The superconducting power divider plays an important role in various superconducting quantum computing circuits and superconducting microwave photon detectors. Therefore, in this paper we investigate how to design and prepare a typical coplanar waveguide superconducting microwave power divider. The parameters are designed by using the odd-even mode method to analyze the transport features of a three-port microwave network. Specifically, the microwave transport properties of the device with a center frequency of 5 GHz and 3 dB power division ratio are simulated. Then, the designed aluminum coplanar waveguide superconducting power divider on silicon is prepared by micro-processing technology and experimentally tested at low temperature. It is shown that the measurement results are consistent with the design parameters. It is noted that the center frequency of the actually prepared power divider is measured to be about 5.25 GHz, which is slightly different from the result of the design and simulation. This difference is probably due to the following main reasons. Firstly, the limited precision of the micromachining process is caused by the fact that the fabricated quarter-wave impedance matching line is etched incompletely, leading the length of the impedance matching line to be shortened. As a consequence, the frequency of the prepared power divider is slightly higher. Secondly, the simulation software is not designed specially for superconducting device simulations, thereby yielding the design parameters slightly different from those of the fabricated superconducting devices. Additionally, a series of attenuations has been used in the experimental test system of the superconducting microwave power dividers for reducing the various noises. This causes the input test signal to weaken, thus the reflected signal turns significantly small. Therefore, none of the S₁₁ parameters of the device can be effectively measured. Finally, neither of S₂₁ and S₃₁ parameters measured in the experiment is the predicted –3 dB, which is mainly due to the imperfections in the welding between SMA connectors and high-frequency transmission lines, and the spot welding between high-frequency transmission lines and power divider samples, and also due to the discontinuities of the high-frequency transmission line and the power divider and so on. All these factors can yield the tested insertion loss of the device. Hopefully, the method in this work can be extended to designing and preparing other passive superconducting microwave devices.