PURPOSE: This study evaluates the relationship between the therapeutic effect of arterial infusion therapy for severe acute pancreatitis and drug distribution on CT-arteriography (CTA). MATERIALS AND METHODS: Eleven patients with severe acute pancreatitis were treated by arterial infusion with use of protease inhibitor and antibiotics. Ten patients had an inflammation of the entire pancreas, while one had pancreatitis localized to the body and tail of the pancreas, The arterial infusion drugs were infused into the celiac artery, splenic artery, inferior pancreaticoduodenal artery, and common hepatic artery, The drug distributions were evaluated by CTA in 10 patients, The duration of arterial infusion ranged from 3 to 39 days, The relationship between the distribution on the CTA and the change in clinical grading of pancreatitis as evaluated by an APACHE II score was studied. RESULTS: Of the nine patients with inflammation of the entire pancreas, six showed the distribution of contrast material to the entire area of pancreatic inflammation (a good distribution) on the CTA, and the remaining three did not show the distribution of contrast material to cover the entire area of pancreatic inflammation (a poor distribution). One patient with localized pancreatitis showed a good distribution, In seven patients with a good distribution, the APACHE II score was decreased from 11.7 points to 4.3 points during follow-up, In the remaining three patients with a poor distribution, the APACHE II score was decreased from 12.3 points to nine points; but was decreased to five points after the additional interventions, One patient without CTA showed a marked improvement in the APACHE II score, No clinically important complications were observed. CONCLUSION The present study findings suggest that arterial infusion is effective in the treatment of severe acute pancreatitis, A good drug distribution to the area of inflammation is needed to ensure a proper therapeutic effect.