1 Previous, in vitro, studies have established the synthesis of platelet activating factor (PAF) by the 're-modelling' pathways in which the activation of a phospholipase A2 (PLA2) enzyme catalyses the hydrolysis of an ether-acyl-phosphocholine to give concomitant release of lyso-PAF, the immediate precursor of PAF, and arachidonic acid, the precursor of the eicosanoids. The aim of this study was to investigate the relationship between PAF and eicosanoid release in human skin, and to study the effect of treatment of skin with a topical steroid, on the release of PAF, lyso-PAF and arachidonic acid. 2 A novel assay procedure was developed for the simultaneous assay of PAF and lyso-PAF in skin exudates from abrasions and suction blisters in normal human skin. In addition we assayed arachidonic acid and prostaglandin E2 (PGE2), a representative eicosanoid. 3 The mean amounts of mediator recovered in the first 30 min period following abrasion were PAF 0.43, lyso-PAF 11.9, PGE2 25.7 and arachidonic acid 760 pmol/sample. The molar ratio of PAF:lyso-PAF: arachidonic acid in skin exudates from abrasions was 1:30:1800 and in suction blister exudates was 1:90:3660. 4 Time course studies showed a decline in the recoveries of arachidonic acid and lyso-PAF, of about 50% in 2 h. In contrast, PAF was recovered in exudates at a constant rate over 2 h but PGE2 release decreased by more than 90% after the initial 30 min period. 5 Topical application under occlusion, of 0.05% clobetasol propionate, a potent corticosteroid, significantly reduced lyso-PAF by 30% in suction blister exudates but did not significantly alter the concentrations of PAF or arachidonic acid. 6 These results show that, in normal human skin, the recovered quantitites of PAF, lyso-PAF and the eicosanoids do not demonstrate a common metabolic pathway for their production. Although corticosteroid treatment reduced lyso-PAF the reduction was not sufficient to affect levels of PAF in skin.