A seed-loading device was designed and modeled using the Monte Carlo method to verify the biological effect of iodine-125 (125I) particles on blood vessels through animal experiments. The dose distribution characteristics of irradiated vessels were established by adjusting the design variables and geometry. The deviation between the actual value and the theoretical value was verified in vitro by the thermoluminescence dosimetry (TLD) method. After verification, the device was used to examine the biological effect of 125I irradiation of canine carotid arteries in two dogs (and one control dog) for 180 days. The hollow cylinder seed-loading device was constructed with an inner diameter of 0.5 cm and a length of 3.3 cm. When six seeds were loaded into a single layer, the source strength ratio of the intermediate layer to the edge layer was 0.7:1. When six layers of seeds were arranged at 0.45-cm intervals, the deviations between the maximum, minimum and mean energy fluence within 2.25 cm of the vessel wall were 2.19% and –4.12%, respectively, and –9% and 4%, respectively, when verified in vitro using TLD. The carotid arteries showed good tolerance to 0.56 kGy (range of 0.51–0.58 kGy) after 180 days of irradiation. In conclusion, this 125I seed-loading device overcomes the random distribution of seeds and lays an accurate radiophysical foundation for subsequent biological experiments. The preliminary results showed that the carotid artery has good tolerance to 0.56 kGy irradiation.