The effect of spinal transections on the preovulatory release of gonadotropins and PRL was investigated in female rats. A preovulatory rise in serum LH, FSH and PRL and subsequent ovulation were prevented by complete spinal transections (CST) at high thoracic levels (T3-T7), but not at low thoracic and lumbar levels (T8-L5), performed at 1000-1230 h on proestrus. Norepinephrine (NE) concentrations in the preoptic-anterior hypothalamic area at 1700-1800 h on proestrus were also significantly reduced by CST at high thoracic levels, but not at lumbar levels. Either electrochemical stimulation of the suprachiasmatic part of the preoptic area or NE injection into the third ventricle at 1400-1500 h on proestrus restored ovulation in animals with CST at high thoracic levels. Animals with CST at lumbar levels exhibited relatively regular 4-day cycles, but those with CST at high thoracic levels showed prolonged periods of diestrous (8-20 days) before they resumed cyclicity. In the case of partial transections, bilateral transections of the lateral columns, but not transections of the dorsal or medial columns, of the spinal cord at T4-T5 significantly blocked the preovulatory gonadotropin release and the occurrence of ovulation. Unilateral transections of the lateral columns of the spinal cord or unilateral electrolytic lesions of the ventrolateral part of the medulla oblongata (VLMO) failed to block ovulation. When combinations of them were performed ipsilaterally, ovulation occurred, but when they were performed contralaterally, the incidence of ovulation was significantly decreased.(ABSTRACT TRUNCATED AT 250 WORDS)