To investigate the secretory pattern of somatostatin (SS) from the median eminence (ME) in the female rat, as well as estrogenic influence on this secretion, we measured both SS release and hypothalamic content in cycling, 10-day ovariectomized, and ovariectomized rats treated with estradiol for 3 days before. Animals were stereotaxically implanted with a push-pull cannula into the ME, and 10 days later the hypothalamic structure was perfused with artificial cerebrospinal fluid for 120-150 min at a regular flow rate of 17 mu l/min. Secretion peaks were observed in the pattern of SS release, whatever the stage of the estrous cycle. The mean amplitude of SS peaks was similar throughout the cycle: 11.7 +/- 4.0, 8.6 +/- 1.5 and 10.5 +/- 1.3 pg at proestrus, estrus and diestrus, respectively, and it was affected neither by ovariectomy (7.4 +/- 1.3 pg) nor by estrogen replacement (5.5 +/- 1.0 pg). By contrast, mean SS release levels in the proestrus phase were significantly higher than those measured in the other phases: 21.6 +/- 2.1 vs. 17.7 +/- 1.2 pg/15 min in diestrus (p < 0.05) and vs. 12.0 +/- 0.7 pg/15 min in estrus (p < 0.001). Hypothalamic SS content showed variations quite similar to those observed during its release, i.e. with the highest values corresponding to the proestrus phase (1,170.5 +/- 224.9 pg/mg of tissue) and to the diestrus (1,156.5 +/- 332.1 pg/mg of tissue) and the lowest values in the estrus (511.5 +/- 52.9 pg/mg of tissue; p < 0.05 vs. proestrus and diestrus). In addition, the lowest SS content and secretion values were found in ovariectomized animals: 95.5 +/- 5.1 pg/mg of tissue (p < 0.001 compared to the values obtained for each stage of the estrous cycle) and 10.0 +/- 0.9 pg/15 min (p < 0.001 vs. proestrus and diestrus), respectively. Patterns of SS release and SS hypothalamic content were not modified by estradiol treatment in ovariectomized animals. Our results suggest that (1) whatever the stage of the estrous cycle, SS release from the ME is not uniform and exhibits irregular peaks; (2) mean SS release levels were subjected to gonadal influence; (3) the occurrence of SS peaks seems to be estrogen-independent, and (4) variations in hypothalamic SS content were generally in good agreement with those of neurohormone release.