Objective To study the dose distribution in tumor target of radioactive esophageal stent covered with 125I seeds which are arranged at different distance. Methods According to the longitudinal and horizontal distance between the 125I seeds, the experiment was divided into group A (0.5 cm), group B (1.0 cm) and group C (1.5 cm). A sheet of white paper, on which a 2cm diameter circle was drawn and was marked with a 5 cm scale, was scanned with a laser scanner, and the scanning data were saved in JPEG format on the computer desktop. By using image conversion program, pictures of 17 layers with the layer space of 5 mm were created. The picture was transmitted to the computer treatment planning system (TPS) to simulate a 2 cm diameter and 8 cm length esophageal stent. Using TPS to delineate the tumor target area (gross tumor volume, GTV), which was used as the simulation of the inner boundary of the tumor target area, outward expansion of 0.5 cm being regarded as the outside border. An annular and 6cm length tube was delineated and it was used as the target area, and the same prescription dose was set, and in each group 125I seeds of 0.3, 0.4, 0.5, 0.6, 0.7, 0.8 and 0.9 mCi were successively loaded on the simulation of esophageal stent. With the help of TPS the DVH graph was obtained, and the D90 (the dose received by 90% of the target) and V90 (the percentage of the GTV receiving 90% the prescription dose) of the three groups were determined. The results were statistically analyzed. Results The arithmetic mean values of D90 of group A, B and C were (217.15± 19.92), (89.16±32.44) and (31.68±11.52) Gy respectively (F=159.18, P<0.05); and the arithmetic mean values of V90 of group A, B and C were (100.00±0.00)%, (86.47±21.36)%and (29.33±21.54)%respectively (F=32.11, P<0.05). Conclusion The 2 cm-diameter radioactive esophageal stent covered with 125I seeds, which have o.6mci activity and are arranged at 1.0cm distance in both longitudinal and horizontal directions, should be recommended in clinical practice.