The excitatory effect of bradykinin (BK) and of low pH on nociceptors appears to partly depend on secondary release of prostaglandins from the surrounding tissue. Rat skin, in vitro, is introduced as a novel model to measure basal and stimulated release of PGE2 and, in future, other substances relevant to nociception, such as neuropeptides. Flaps of hairy skin (n=57) from the rat saphenous region of the hindpaw were subcutaneously excised and fixed on acrylic rods, the corium side exposed. The preparations were equilibrated in carbogen gassed "synthetic interstitial fluid" (SIF) for 30 minutes. The skin flaps were then immersed for 5 minutes each in 9 consecutive glass tubes, which were mounted in a shaking bath at 32 degrees C. Each tube was filled with 5 ml of gassed SIF, the third tube contained inflammatory mediator(s) dissolved in SIF or solutions of low pH. After passage of the skin flap, the eluates were deep frozen (-70 degrees C) and the PGE2 content measured, off-line, using an enzyme immuno-assay. As stimulants, BK at 10(-5) M (n=9) and 10(-6) M (n=4) and BK in equimolar combination with histamine (HA) and serotonin (5-HT; 10(-5) M: n=8, 10(-6) M: n=6, 10(-7) M: n=6) dose-dependently increased PGE2 release. Considering the total amount of PGE2 secreted the combination of inflammatory mediators caused a significantly greater release of PGE2 at 10(-5) and 10(-6) M (p<0.01, Kruskal-Wallis test) than BK stimulation alone. Racemic flurbiprofen caused a profound depression of basal and stimulated release. Solutions of high proton concentration are known to stimulate and sensitize nociceptors. However, phosphate buffered SIF at pH 6.1 and 6.4 caused a substantial and significant decrease of the PGE2 release, probably due to low-pH block of phospholipases. Thus, algogenic potency of mediators does not necessarily match their pro-inflammatory action.