To decrease the toxicity of hepatic arterial fluorodeoxyuridine (FUDR) administered through an Infusaid pump (Shiley Infusaid, Inc., Norwood, MA), 50 patients with liver metastases from colorectal cancer were selected randomly to receive FUDR, 0.3 mg/kg/d, for 14 of 28 days, with or without a total dose of 20 mg of hepatic arterial dexamethasone for 14 of 28 days. Patients were stratified according to the percentage of liver involvement by tumor and the perfusion pattern on macroaggrated albumin perfusion scan (MAA) scan. There was a trend toward decreased frequency of bilirubin levels in the group receiving dexamethasone plus FUDR versus the group receiving FUDR alone (9% and 30%, respectively, had a 200% or greater increase from baseline; P = 0.07). Patients in the group treated with dexamethasone and FUDR received higher doses of FUDR in the second, third, fifth, and sixth months than those receiving FUDR alone; however, this was statistically significant only in the fifth month (percentages of planned dose received: 42% and 19%, respectively; P = 0.05), and there was no overall difference for the total 6-month period. The complete and partial response rates were increased in patients receiving dexamethasone and FUDR versus FUDR alone (8% and 63% versus 4% and 36%, respectively; P = 0.03), and there was a trend toward increased survival with the addition of dexamethasone (median, 23 months and 15 months, respectively; P = 0.06). In conclusion, the use of hepatic arterial dexamethasone is associated with an increased response rate and a trend toward increased survival and decreased bilirubin levels. Therefore, the authors recommend additional investigation of the use of dexamethasone with chemotherapy to treat hepatic metastases.