1. In the guinea-pig isolated vas deferens preparation bathed in Tyrode's solution, the prostacyclin analogues, cicaprost, TEI-9063, iloprost, taprostene and benzodioxane-prostacyclin, enhanced twitch responses to submaximal electrical field stimulation (20%-EFS). The high potency of cicaprost (EC150 = 1.3 nM) and the relative potencies of the analogues (equi-effective molar ratios = 1.0, 0.85, 1.6, 17 and 82, respectively) suggest the involvement of a prostacyclin (IP-) receptor. 2. Maximum enhancement induced by cicaprost in 2.5 mM K+ Krebs-Henseleit solution was similar to that in Tyrode solution (2.7 mM K+), but was progressively reduced as the K+ concentration was increased to 3.9, 5.9 and 11.9 mM. There was also a greater tendency for the other prostacyclin analogues to inhibit EFS responses in 5.9 mM standard K+ Krebs-Henseleit solution; this may be attributed to their agonist actions on presynaptic EP3-receptors resulting in inhibition of transmitter release. 3. The EFS enhancing action of cicaprost was not affected by the alpha1-adrenoceptor antagonist prazosin (100 and 1000 nM). Cicaprost (20 and 200 nM) did not affect contractile responses of the vas deferens to either ATP (5 microM) or alpha,beta-methylene ATP (1 microM) in the presence of tetrodotoxin (TTX, 100 nM). In addition, enhancement by cicaprost of responses to higher concentrations of ATP (30 and 300 microM) in the absence of TTX, as shown previously by others, was not seen. Prostaglandin E2 (PGE2, 10 nM) and another prostacyclin analogue TEI-3356 (20 nM) enhanced purinoceptor agonist responses. Unexpectedly, TTX (0.1 and 1 microM) partially inhibited contractions elicited by 10-1000 microM ATP; contractions elicited by 1-3 microM ATP were unaffected. Further studies are required to establish whether a pre- or post-synaptic mechanism is involved. 4. In a separate series of experiments, cicaprost (5-250 nM), TEI-9063 (3-300 nM), 4-aminopyridine (10-100 microM) and tetraethylammonium (100-1000 microM) enhanced both 20%-EFS responses and the accompanying overflow of noradrenaline to a similar extent. In further experiments with the EP1-receptor antagonist AH 6809, TEI-3356 (1.0-100 nM) and the EP3-receptor agonist, sulprostone (0.1-1.0 nM) inhibited both maximal EFS responses and noradrenaline overflow, thus confirming previous reports of the high activity of TEI-3356 at the EP3-receptor. Cicaprost had no significant effect on noradrenaline overflow at 10 and 100 nM, but produced a modest inhibition at 640 nM. 5. In conclusion, our studies show that prostacyclin analogues (particularly TEI-3356) can inhibit EFS responses of the guinea-pig vas deferens by acting as agonists at presynaptic EP3-receptors. Prostacyclin analogues (particularly cicaprost and TEI-9063) can also enhance EFS responses through activation of IP-receptors. The mechanism of the enhancement has not been rigorously established but from our results we favour a presynaptic action to increase transmitter release.