Abstract: This study aims to promote the intelligent development of field crop sowing for low labor intensity. A front and rear dual-drive and dual-steering robot chassis was designed for maize sowing using a wireless remote control system. Some components were equipped to realize the unmanned maize sowing. The overall structure and layout of the chassis were also determined to fully meet the requirements of maize sowing in the maize stubble fields. The key components were designed as well, such as the frame, drive steering system, arch leveling mechanism, and control system. According to theoretical calculation, two AC asynchronous motors of 2.2 kW were used to provide the driving force for the chassis. A column-type electric power steering was used to obtain the front and rear wheel steering. A wire-controlled brake was used for the remote braking during sowing. The vehicle control system was constructed with the VCU (vehicle control unit) as the core controller. Radio signals were also transmitted between the remote control and VCU. CAN bus communication was adopted between VCU and the rest subsystems. The static and modal analysis was performed on the frame using the finite element method, in order to verify the structural design and material strength of the frame. As such, the maximum equivalent stress of the frame was 98.2 MPa at the constant speed under full load bending. The maximum deformation was 1.84 mm. In terms of full load torsion, the maximum equivalent stress of the frame was 106.65 MPa at the constant speed, and the maximum deformation was 2.05 mm. The stress at each position of the frame was less than the standard stress of the material. The constrained modal analysis showed that the first-order natural frequency of the frame was 34.442 Hz, and the maximum deformation was 6.415 9 mm. While the vibration frequencies caused by terrain undulations and ditching were generally less than 20 and 10 Hz, respectively. Therefore, the influence of resonance on the performance of the sowing components was effectively avoided after optimization. The finite element simulation showed that the structural strength of the frame fully met the requirements of maize sowing, indicating a reasonable design. The synchronous reverse steering of the front and rear wheels was adopted to reduce the turning radius for the high efficiency of the chassis. Theoretical analysis showed that the smallest turning radius of the chassis was achieved when the front and rear wheels were rotated in opposite on the same side that deflect by the same angle. Specifically, the turning radius values of the inner and outer wheels were 3.44 and 4.65 m, respectively. Once the chassis was traveling on a slope, the force analysis showed that the maximum overturning angles for the uphill and downhill slopes were 69° and 67.7°, respectively, while the lateral overturning angle and the lateral slip angle were 58.3° and 33.8°, respectively. The performance test of the prototype showed that the chassis safely and stably went through 8°-15° slopes and overcame 6-15 cm obstacles.It infers that the chassis shared the better performance of climbing stability and obstacle crossing. The maize planting was successfully realized under various complex surfaces. The better driving performance of the chassis was achieved with the average driving speeds of 1.461 m/s on corn stubble fields, with an average straight-line deviation rate of 4.42%. The qualified rate of the sowing depth was greater than 86% in the chassis of the maize sowing robot under different test conditions. Nevertheless, the qualified rate of sowing depth decreased slightly with the increase in the operating speed. There was a relatively small variation in the standard deviation and the coefficient of variation of sowing depth. The better stability of the chassis was achieved in the sowing depth under different speeds and soil compaction. The average spacing of qualified seed was 24.32 cm, and the coefficient of variation of seed spacing was 5.1%. There was no re-seeding or missed sowing. The field vibration and excitation shared no influence on the working performance of the seed meter. The average row spacing was 69.3 cm, while the standard and maximum deviation of row spacing were 2.27 and no more than 4 cm, respectively. The chassis can be expected to achieve uniform row spacing during sowing.