The aim of the present study was to investigate the mechanism of the retrograde transfer of ovarian steroid hormones from the ovarian lymphatic and venous effluent to the arterial blood supplying the ovary. In the first experiment, reproductive organs were collected from gilts in the luteal (n = 10) and follicular (n = 10) phase of the oestrous cycle. The ovary with the mesovarium was isolated and perfused through the ovarian artery with warmed, oxygenated autologous blood. The concentrations of progesterone and oestradiol in ovarian arterial blood increased on passing through the ovarian artery to the ovary, in the luteal phase, from 20.3 +/- 2.1 to 31.4 +/- 3.9 ng ml(-1) (P < 0.001) and from 6.2 +/- 0.8 to 11.4 +/- 1.4 pg ml(-1) (P < 0.001), respectively, and in the follicular phase, from 1.2 +/- 0.2 to 2.2 +/- 0.4 ng ml(-1) (P < 0.001) and from 8.2 +/- 1.8 to 13.2 +/- 2.3 pg ml(-1) (P < 0.001), respectively. Approximately 17.5 +/- 3.9 % of the progesterone and 12.6 +/- 1.7 % of the oestradiol found in the ovarian venous effluent was retrogradely transferred from the ovarian venous blood to the ovary in the luteal phase. In the follicular phase, these values were 10.1 +/- 2.0 % and 8.6 +/- 1.4 %, respectively. The efficiency of retrograde transfer of oestradiol and the rate of retrograde transfer of progesterone differed between phases of the oestrous cycle (P < 0.05 and P < 0.0001, respectively). A direct relationship between the concentration of the steroids in the venous effluent and the efficiency and rate of the retrograde transfer to the ovary was not found. In the second experiment (luteal phase, n = 10; follicular phase, n = 5), the concentration of progesterone and oestradiol increased in both ovarian arterial blood (P < 0.0001) and in the venous effluent (P < 0.0001) after administration of the steroids into the lymphatic vessels of the isolated mesovarium with separated ovary. In the third experiment (follicular phase, n = 5), with the mesovarium isolated after the ovary was removed and ovarian venous blood flowing out under the force of gravity (without the blood pressure in the ovarian vein), it was demonstrated that the veno-venous network covering the branches of the ovarian artery was supplied with the blood flowing out from the mesovarian tissue and that the filling of the veno-venous network was dependent on the blood pressure in the ovarian artery. We conclude that the effective retrograde transfer of steroid hormones from ovarian venous and lymphatic effluent to the ovary is accomplished not only by the classical counter-current exchange mechanism, but also as a result of complex processes that may be dependent on a specific part of the circulation of the blood and lymph in the periovarian vascular complex of the mesovarium.