1. Neurointermediate lobes of rat pituitaries were incubated in Locke or Krebs solution, and the vasopressin released into the medium was assayed on the blood pressure of the pithed rat or by a radioimmunological procedure. Release of vasopressin over resting levels was evoked either by incubation with 60 mM KCl (high K) solution or by electrical stimulation of the pituitary stalk. Two different kinds of electrical stimulation were carried out. Procedure A (1 ms, 10 Hz, 5 times for 1 min within 10 min) induced a vasopressin overflow which was greatly calcium-dependent but only insignificantly sensitive to tetrodotoxin (TTX). Procedure B (0.2 ms, 15 Hz, 10 s trains with 10 s intervals for 10 min) evoked a completely calcium-dependent and TTX-sensitive vasopressin overflow. 2. The vasopressin output evoked by high K was unaltered in the presence of dopamine 10 μM or apomorphine 100–300 μM. 3. The vasopressin overflow evoked by stimulation procedure A was decreased by 50–67% in the presence of apomorphine 10 μM. However, the dopamine antagonists sulpiride 1–100 nM and flupenthixol 10 μM also inhibited the vasopressin release by 30–40%. Nevertheless, the combination of apomorphine 10 –M with sulpiride 1 nM or flupenthixol 10 μM resulted in a significant increase of the vasopressin release if compared with apomorphine (or sulpiride) alone. Thus, apomorphine antagonized the inhibition of the vasopressin release by the dopamine antagonists, and these drugs blocked the inhibitory action of apomorphine, indicating complex but specific agonist-antagonist interactions. Naloxone 1 μM which failed to change vasopression overflow, prevented the inhibitory effect of sulpiride 100 nM. This indicates that the inhibition of vasopressin release by sulpiride may be mediated by an increased secretion of endogenous opioids probably from the intermediate lobe. 4. The vasopressin overflow evoked by stimulation procedure B was decreased by 20% in the presence of apomorphine 100 nM or 1 μM. Bromocriptine had a biphasic effect on vasopressin release, stimulating it by 20% at 1 nM and inhibiting it by 35% at 3 μM. Sulpiride 10 nM to 1 μM stimulated the vasopressin release by about 20% and antagonized the inhibitory actions of apomorphine and bromocriptine in a concentration-dependent manner. However, sulpiride 100 nM had no effect on the facilitatory action of bromocriptine 1 nM. 5. These findings suggest a complex dopaminergic modulation, in the neurointermediate lobe, of vasopressin release evoked by intermittent electrical stimulation, and support a physiological role for the intrinsic dopaminergic fibres described by others. We propose the existence of two dopamine receptors, one mediating inhibition of vasopressin release (activated by apomorphine, bromocriptine at high concentrations or endogenous dopamine, and blocked by sulpiride), and the other one mediating facilitation of release (activated by endogenous dopamine, and blocked by flupenthixol). It is further suggested that under the field-stimulation conditions of procedure A an endogenous opioid is secreted which inhibits vasopressin release and which itself is under inhibitory control by endogenous dopamine.