1. The prostanoid receptor(s) that mediates inhibition of bacterial lipopolysaccharide (LPS)-induced tumour necrosis factor-alpha (TNF-alpha) generation from human peripheral blood monocytes was classified by use of naturally occurring and synthetic prostanoid agonists and antagonists. 2. In human monocytes that were adherent to plastic, neither prostaglandin D2 (PGD2), prostaglandin E2 (PGE2), prostaglandin F(2 alpha) (PGF(2 alpha)) nor the stable prostacyclin and thromboxane mimetics, cicaprost and U-46619, respectively, promoted the elaboration of TNF alpha-like immunoreactivity, as assessed with a specific ELISA, indicating the absence of excitatory prostanoid receptors on these cells. 3. Exposure of human monocytes to LPS (3 ng ml-1, approximately EC84) resulted in a time-dependent elaboration to TNF alpha which was suppressed in cells pretreated with prostaglandin E1 (PGe1), PGE2 and cicaprost. This effect was concentration-dependent with mean pIC50 values of 7.14, 7.34 and 8.00 for PGE1, PGE2 and cicaprost, respectively. PGD2, PGF(2 alpha) and U-46619 failed to inhibit the generation of TNF alpha at concentrations up to 10 microM. 4. With respect to PGE2, the EP-receptor agonists, 16,16-dimethyl PGE2 (non-selective), misoprostol (EP2/EP3-selective), 11-deoxy PGE1 (EP2-selective) and butaprost (EP2-selective) were essentially full agonists as inhibitors of LPS-induced TNF alpha generation with mean pIC50 values of 6.21, 6.02, 5.67 and 5.59, respectively. In contrast to the results obtained with butaprost and 11-deoxy PGE1, another EP2-selective agonist, AH 13205, inhibited TNF alpha generation by only 21% at the highest concentration (10 microM) examined. EP-receptor agonists which have selectively for the EP1- (17-phenyl-omega-trinor PGE2) and EP3-receptor (MB 28,767, sulprostone) were inactive or only weakly active as inhibitors of TNF alpha generation. 5. Pretreatment of human monocytes with the TP/EP4-receptor antagonist, AH 23848B, at 10, 30 and 100 microM suppressed LPS-induced TNF alpha generation by 10%, 28% and 77%, respectively, but failed to shift significantly the location of the PGE2 concentration-response curves. 6. Given that AH 13205 was a poor inhibitor of TNF alpha generation, studies were performed to determine if it was a partial agonist and whether it could antagonize the inhibitory effect of PGE2. Pretreatment of human monocytes with 10 and 30 microM AH 13205 inhibited the generation of TNF alpha by 31% and 53%, respectively, but failed to shift significantly the location of the PGE2 concentration-response curves at either concentration examined. 7. Since PGD2 and 17-phenyl-omega-trinor PGE2 (EP1-agonist) did not suppress TNF alpha generation, the EP1/EP2/DP-receptor antagonist, AH 6809, was employed to assess if EP2-receptors mediated the inhibitory effect of PGE2. Pretreatment of human monocytes with 10 microM AH 6809 did not affect LPS-induced TNF alpha generation but produced a parallel 3.5 fold rightwards shift of the PGE2 concentration-response curve. 8. Collectively, these data suggest that human peripheral blood monocytes express at least two distinct populations of inhibitory prostanoid receptors that mediate inhibition of LPS-induced TNF alpha generation. One of these probably represents i.p. receptors based upon the selectivity of cicaprost for this subtype. The other population has the pharmacology of EP-receptors, but the rank of potency for a range of synthetic EP-receptor agonists was inconsistent with an interaction with any of the currently defined subtypes. Given the pharmacological behaviour of butaprost, AH 6809 and AH 23848B in these cells, we propose that multiple (EP2- and/or EP-4- and/or i.p.) or novel EP-receptors mediate the inhibitory effect of PGE2 on TNF alpha generation.