Higher response rates in colorectal cancer have been observed with regimens that increase the cytotoxicity of fluorouracil (5-FU) by altering the biochemical milieu at its site(s) of action. Phosphonacetyl-L-aspartate (PALA), which inhibits aspartate transcarbamylase and depletes uridine nucleotide pools in vitro and in vivo, selectively potentiates the antitumor activity of 5-FU in preclinical models. In a phase I/II study in patients with advanced colorectal cancer, PALA 250 mg/m2 was given on day 1, followed 24 hours later by 5-FU 2,600 mg/m2 by 24-hour infusion repeated weekly. Through the use of subcutaneous ports and portable infusion pumps, all patients were treated outside the hospital setting. Thirty-nine patients without prior chemotherapy received 884 courses of treatment. The primary site was colon in 31, rectum in eight. Toxicity was generally mild to moderate except among four patients who were escalated to 5-FU 3,250 mg/m2: two developed severe gastrointestinal toxicity and myelosuppression. Among the remaining 35 patients, gastrointestinal and neurologic toxicities predominated, but usually did not develop until the third or fourth month of treatment. Two patients were inevaluable for response. Among the 37 evaluable patients there were three complete and 13 partial remissions for a total response rate of 43% (95% confidence interval [Cl], 27% to 59%). The median duration of response was 5 months (range, 1.5 to 15 months). The projected mean survival is in excess of 17 months. This high response rate is comparable to the best results obtained with other means of modulation of 5-FU. Demonstration of a survival benefit will require larger phase III studies.