The objectives of this phase I study were to determine the maximum-tolerated dose (MTD), dose-limiting toxicities (DLTs), and preliminary efficacy of intraperitoneally administered irinotecan (CPT-11) in gastric cancer patients with peritoneal seeding. Gastric adenocarcinoma patients with surgical biopsy proven peritoneal seeding were enrolled at the time of surgery. Prior to IP chemotherapy, patients underwent palliative gastrectomy and CAPD catheter insertion in which CPT-11 was administered on postoperative day 1. The IP CPT-11 was initiated at 50 mg/m(2), which was escalated to 100, 150, 200, 250, and 300 mg/m(2). IP CPT-11 chemotherapy was repeated every 3 weeks. Seventeen patients received a total of 56 cycles at five different CPT-11 dose levels. The DLTs were neutropenic fever, neutropenia, and diarrhea. At the dose level 2 (100 mg/m(2)), there were one DLTs in one of the first cohort of three patients, but no DLTs at the second cohort of this level. At the dose level 5 (250 mg/m(2)), two DLTs were detected in the first two patients; thus, the accrual was stopped resulting in the recommended dose of IP CPT-11 of 200 mg/m(2). Median progression-free survival was 8.6 months (95% CI, 5.9,11.2), and median overall survival was 15.6 months (95% CI, 8.4,22.8). Pharmacokinetic results of the study showed that the C (max) of peritoneal SN-38 was achieved earlier than that of plasma SN-38. Intraperitoneally administered CPT-11 was feasible and tolerable. Further, phase II study of IP CPT-11 in gastric cancer patients with peritoneal seeding is warranted.