Objective To compare the rates of entry and intake of commensal rodents for poison bait boxes made of concrete and ceramic in Shenzhen, China, and the influence of different poison baits on the rates of entry and intake of commensal rodents, and to provide a scientific basis for the purchase of poison bait boxes and poison baits for deratization in Shenzhen. Methods In July 2018, the villages of Shuiku Xincun in Luohu district and Shawei in Futian district were selected as experimental points. Forty-nine groups of poison bait boxes were placed at each experimental point. Each group consisted of one concrete box and one ceramic box. The two boxes in each group were laid 0.5 m apart according to the specifications, while the space between each two groups was 5 m. For each box, poison bait was added and replaced every day, maintaining 30 g of bromadiolone poison grains or 2 pieces of bromadiolone wax bait (15 g each). We observed and recorded regularly every day, for 9 days in Shuiku Xincun and 10 days in Shawei. Statistical analysis of the recorded data was performed using chi-square test. Results The mean rates of entry for concrete and ceramic poison bait boxes were 35.88% and 24.70%, respectively, showing a significant difference (χ²=27.510, P<0.001), while the mean rates of intake for concrete and ceramic poison bait boxes were 17.08% and 6.87%, respectively, also showing a significant difference (χ²=45.977, P<0.001). The mean rates of entry for boxes containing bromadiolone poison grains and bromadiolone wax bait were 7.55% and 11.84%, respectively, giving a significant difference (χ²=5.140, P=0.023), while the mean rates of intake for boxes containing the above two baits were 5.51% and 9.59%, respectively, also giving significant difference (χ²=5.847, P=0.016). The mean rates of entry and intake of commensal rodents for poison bait boxes in different placement positions were 53.33% vs 35.14% (χ²=12.547, P<0.001), and 26.67% vs 13.78% (χ²=10.689, P=0.001), respectively. In this experiment, the rates of entry and intake for the concrete poison bait box were better than those for the ceramic poison bait box, while the rates of entry and intake for boxes containing bromadiolone wax bait were better than those for boxes containing bromadiolone poison grains. In addition, the poison bait boxes near the rodent hole had higher rates of entry and intake than those not close to the hole. Conclusion In this experiment, when the outdoor poison bait was used in the residential area of the experimental area, the acceptability of the concrete poison bait box to commensal rodents and the rate of intake of the poison bait were significantly better than those for the ceramic poison bait box. The commensal rodents preferred the poison bait boxes containing bromadiolone wax bait during feeding process. Concrete poison bait boxes and bromadiolone wax bait should be selected as a priority in the rodent control activities in Shenzhen.