(1) Output of acetylcholine (ACh) from fetal vessels of Krebs-perfused human placental cotyleda was estimated by bioassay using the rat stomach strip. (2) Infusion of high concentrations of the opioids pethidine, morphine or ethylketocyclazocine, but not [D-Ala2]-methionine enkephalinamide (0.1-300 mumol/l), reduced ACh output. Fifty per cent inhibition of output occurred in the presence of concentrations of greater than or equal to 300 mumol/l (morphine and ethylketocyclazocine) and 338 (317, 353; 95 per cent c.l., n = 6) mumol/l (pethidine). (3) ACh output was inhibited by infusion of 100 mumol/l naloxone or 10 mumol/l naltrexone, but was not affected by lower concentrations of either antagonist. (4) These results suggest that the therapeutic concentrations of morphine and pethidine likely to occur in vivo would not affect placental ACh output into fetal vessels. The finding that high concentrations of synthetic opioids or opioid antagonists were required to inhibit output suggests that they may not be acting specifically, and provides no evidence for the hypothesis that endogenous opioids play a role in control of ACh release into fetal vessels of human placentae.