Thirty-two percent dextran 70 was administered to 53 patients receiving intraperitoneal (i.p.) chemotherapy in an attempt to better maintain catheter function. One hundred milliliters of 32% dextran 70 was administered i.p. at the time of catheter placement and at the completion of each course of chemotherapy (every 3 to 4 weeks). Analysis of the functional survival of the dextran treated catheters and 20 historical controls was performed. The cumulative probabilities of catheters maintaining bi-directional function in the dextran treated and control groups were 0.75 and 0.50 respectively. This difference was statistically significant at p = 0.051 by two-tailed Wilcoxon analysis. The difference between survival of dextran treated and control catheters increased if patients who received intraperitoneal doxorubicin were factored out (p = 0.035 by two-tailed Wilcoxon analysis). Plasma and peritoneal dextran levels were measured on 9 courses in 8 patients. Dextran was detectable in the peritoneal cavity up to 7 days after administration. The "apparent half-life" of dextran 70 in the peritoneal cavity was 36 hours. Plasma dextran concentrations increased for 2 days following i.p. administration and then decreased with an apparent half-life of 36 hours. One patient experienced chills and another had an anaphylactoid reaction following administration of the dextran. This study suggests that i.p. administration of 32% dextran 70 may be an effective means of minimizing peritoneal catheter failures.