1. Interactions between indomethacin, noradrenaline and vasodilators were studied in rings of ductus arteriosus isolated from fetal rabbits. The effect of incubation with prostaglandin E2 (PGE2) on the noradrenaline concentration-contraction response curve was studied in the presence and absence of indomethacin. Also, the ductus was pre-contracted with 10 microM noradrenaline and concentration-relaxation response curves (CRRC) to PGE2, cicaprost, cromakalim and forskolin were obtained in the presence and absence of indomethacin. 2. In the absence of indomethacin, PGE2 (from 1 nM to 100 nM) decreased the pEC50 to noradrenaline to a maximum of 0.4 to 0.5 log units (i.e. an approximately three fold increase in EC50 [M]). In the presence of 1 microM indomethacin, PGE2 (0.1 nM to 100 nM) decreased the pEC50 to noradrenaline by 2.39 log units (i.e. a 245 fold increase in EC50). By comparing the control pEC50 to noradrenaline with the relationship between the pEC50 to noradrenaline and [PGE2] in 1 microM indomethacin, the effect of endogenous PGE2 synthesized in the vessel wall was estimated as being equivalent to a bath concentration of approximately 1 nM exogenous PGE2. 3. When the vessel was pre-contracted with 10 microM noradrenaline, indomethacin had no effect on the CRRC to PGE2 but did alter the CRRC to other vasodilators. The sensitivity of the vessel to cicaprost, cromakalim and forskolin was decreased in 1 microM indomethacin compared with control. Forskolin caused complete relaxation in the presence and absence of indomethacin. Indomethacin decreased the maximum response to cromakalim but increased the maximum response to cicaprost. PGE2, 0.3 nm, partially reversed the effect of indomethacin on the sensitivity of the vessel to forskolin.4. We conclude that under varying experimental conditions, indomethacin increased the sensitivity of the ductus to the effects of PGE2 but decreased its sensitivity to other vasodilators. Both effects can be explained by elimination of endogenous PGE2. However, indomethacin increased the maximum response to cicaprost, which cannot be explained by elimination of endogenous PGE2 but may be due to elimination of endogenous prostacyclin.