The peritoneal plasma barrier (PPB) is a pharmacologic entity of importance for treatment planning in patients with malignant tumors confined to the abdominal cavity. We have examined the pharmacokinetics of the PPB by sampling abdominal fluid following intravenous mitomycin C (MMC) administration. The study included 15 cycles of treatment in seven patients with peritoneal carcinomatosis from colorectal cancer. Five patients were studied twice and one patient was studied three times for a total of 15 cycles. Patients were treated with intraperitoneal 5-fluorouracil (5-FU) at 20 mg/m2 in 11 of fluid. Between 250 and 500 ml of ascites remained after the 23 hour intraperitoneal dwell. On day 3, MMC (12 mg/m2) was administered intravenously as a 2-hour continuous infusion in 200 ml of dextrose solution. The concentration of MMC was determined in plasma, peritoneal fluid, and urine by high performance liquid chromatolography (HPLC) at frequent intervals for 8 hours. The area under the curve (AUC) for plasma as related to peritoneal fluid was three times greater for plasma in one cycle, two times greater for plasma in three cycles, 1.5 times greater for plasma in five cycles, and the same in six cycles. AUC ratios showed a correlation with the extent of peritoneal stripping at the prior surgical procedure 6 weeks to 14 weeks previously. We conclude that malignant ascites may be less exposed to chemotherapy than systemic tumor nodules when the intravenous route of drug administration is used. This inadequacy is even more pronounced in patients who have had extensive abdominal surgery.