Resveratrol (Res) inhibits ovarian cancer (OC) cell growth but its in vivo anti-OC effects are unclear due to the low bioavailability of systemically administered Res. Intraperitoneal administration may overcome this therapeutic dilemma because it makes Res directly affect the abdominal tumors. Ethanol and DMSO are common Res solvents, while their reliability and safety for long-term in vivo treatment remain unknown. A rat orthotopic OC model was established using the rat NUTU-19 OC cell line. Res dissolved in 10% ethanol or 0.2% DMSO was injected intraperitoneally (20 mg/kg/day) into tumor-free and tumor-bearing rats for 2 weeks. The tumors were collected for gross, morphological and molecular examinations, and blood and ascitic samples were obtained for a CA125 ELISA. Res concentration in ovarian tissues was determined by high performance liquid chromatography (HPLC). The average tumor weight (0.187±0.065 g) of the Res-in-DMSO group was lower than that of untreated (0.426±0.091 g; P<0.01) and Res-in-ethanol (0.238±0.073 g; P<0.05) group. The average bloody ascitic volumes collected from untreated, Res-in-ethanol, and Res-in-DMSO groups were 5.65±0.27, 2.75±0.14, and 2.09±0.11 ml, respectively. Abundant TUNEL-positive cells, ARHI and PIAS3 upregulation, CA125 reduction, and decreased STAT3 nuclear translocation were found in the Res-in-ethanol and, especially, the Res-in-DMSO group. Widespread plaques of Res deposits were found on the abdominal serosa of the Res-in-ethanol group, but not in the Res-in-DMSO group. HPLC revealed a higher Res concentration in Res-in-DMSO-treated tumor tissues than in those treated by Res-in-ethanol (P<0.01). Fertility was maintained after long-term Res treatment. Intraperitoneal administration of Res effectively inhibited rat orthotopic ovarian cancer growth without affecting normal tissues. The Res-in-DMSO group had the highest drug bioavailability and therefore stronger tumor-suppressive effects on ovarian cancer tissues.