We used our nonsurgical technique for collecting pituitary venous blood to relate the dynamics of thyrotrophin-releasing hormone (TRH) secretion to the secretion patterns of both prolactin and thyrotrophin in periovulatory mares, either euthyroid (n = 5) or made hypothyroid by treatment with propyl-thiouracil (n = 5). Pituitary venous blood was collected continuously and divided into 1-min aliquots for 4 h. To test the effect of dopamine on the relationship between secretion patterns, sulpiride, a selective D2 receptor antagonist, was given i.m. after 2 h of sampling. Thorough testing of the model and blood collection procedure revealed no sites of TRH loss. Hypothyroidism increased the mean secretion rates of TRH (P = 0.04) and thyrotrophin (P < 0.0001) but not prolactin. Sulpiride increased prolactin secretion rates in hypothyroid (P < 0.0001) and control (P = 0.007) mares, but did not alter TRH or thyrotrophin secretion rates. In both groups of mares, all three hormones were secreted episodically but not rhythmically. In both groups, the secretion pattern of TRH was almost always significantly related to that of thyrotrophin, as assessed by cross correlation and cross approximate entropy (ApEn) analysis. However, the degree of linear correlation was weak, with only 14% (hypothyroid) or 8% (controls) of the variation in thyrotrophin secretion rates attributable to TRH. Prolactin and TRH secretion patterns before sulpiride were coupled on cross ApEn analysis in both groups, and the minute-to-minute secretion rates of the two hormones were correlated in four hypothyroid and three euthyroid mares. Overall, the small, but significant, degree of association between TRH and prolactin was similar to that between TRH and thyrotrophin. In hypothyroid mares, sulpiride increased (P = 0.02) the synchrony between TRH and prolactin patterns. We conclude that in horses: (i) little TRH degradation occurs during passage through the pituitary or in blood after 1 h at 37 degrees C; (ii) TRH is not the major factor controlling minute-to-minute fluctuations in either thyrotrophin or prolactin; and (iii) reducing two strongly inhibitory inputs (i.e. dopamine and thyroid hormones) may magnify the stimulatory effect of TRH on prolactin secretion.