Objective To investigate the delivery efficacy of periadventitial delivery of 125I-iododeoxyuridine (125I-IUdR) in comparison of intravascular delivery to determine the optimal delivery method for inhibiting post-angioplasty restenosis. Methods In 8 pigs, one side carotid, subclavian and iliac arteries of each pig were injured by balloon angioplasty with a 20% overstretches. Then, 4 mCi of 125I-IUdR was delivered at each targeted vessel with periadventitial method in 4 pigs (periadventitial group) and with intravascular method via a porous balloon catheter in other 4 pigs (intravascular group). The animals survived for 5 hours and the blood radioactivity was investigated prior to and hourly after procedure until sacrifice. The targeted vessels and renal arteries (for control) were harvested for gamma-counting and histological observation. Meanwhile, the radioactivity in thyroid, liver, bladder, small bowel and each kidney also were measured to determine the biodistribution of 125I. The activities of 125I presented in arterial and tissue specimens were compared between the two delivery groups. The targeted arteries were histologically observed and the ratio of intima to media (I∶M ratio) was calculated. Results The target arterial walls in the periadventitial group had 3.4 times as much of 125I radioactivity as in the intravascular group, respectively (P=0.038); the blood activity in intravascular group was significantly higher than periadventitial group immediate after procedure (P<0.05) and intravascular delivery resulted in much higher activity in urine than periadventitial delivery (P<0.05). The systemic biodistributions of 125I-IUdR in the organs were slightly higher in the intravascular group (P>0.05). The mean I: M ratios in both groups were 0.05 without additional injury at the vessel wall. Conclusion The periadventitial delivery offered substantial advantage over intravascular approach with high local delivery efficacy. The apparent redistribution rate is more rapid following intravascular delivery.