1. When injected into a 6-day-old mouse air-pouch, human recombinant interleukin-8 (IL-8; 0.03-3 micrograms) induced, in a dose-dependent fashion, an accumulation of neutrophils which could be reliably assessed 4 h after the injection. No protein extravasation was measured above the values obtained with the vehicle alone (carboxymethylcellulose, CMC, 0.5% w/v in phosphate-buffered solution, PBS). 2. The IL-8 effect (routinely evaluated at 1 microgram dose) was inhibited neither by local administration of actinomycin D (1 microgram) nor by systemic treatment with indomethacin (1 mg kg-1, i.v.), BWA4C (5 mg kg-1, p.o.), methysergide (6 mg kg-1, i.p.) and RP67580 (2 mg kg-1, i.p.). 3. Treatment of mice with the H1 antagonist, mepyramine (1-10 mg kg-1, i.p.) resulted in a dose-dependent inhibition of the cell accumulation elicited by the chemokine, with a maximal reduction of approximately 50-60%. The mepyramine effect was not due to a non specific reduction of neutrophil function, since treatment with this drug (6 mg kg-1, i.p.) did not modify the cell infiltration measured in response to a challenge with interleukin-1 beta (20 ng) or with the vehicle CMC to any extent. Moreover, treatment of mice with mepyramine did not modify cell counts in a peripheral blood film with respect to controls. Two other H1 antagonists, chemically unrelated to mepyramine, diphenhydramine (9 mg kg-1, i.p.) and triprolidine (0.5 mg kg-1, i.p.), inhibited IL-8-induced migration to a similar extent (approximately 50-60%), whereas the H2 antagonist, ranitidine (5 mg kg-1, i.p.) was without effect. 4. The concept that endogenous histamine could be involved in the IL-8 effect was strengthened in two ways: (i) addition of histamine (0.2-2 microg) to a small dose of IL-8 (0.3 microg) potentiated the cell elicitation induced by the chemokine without having any effect on its own; (ii) IL-8-induced neutrophil accumulation was greatly impaired in animals depleted of mast cell amines by sub-chronic (5 day) treatment with compound 48/80 according to an established protocol.5. The glucocorticoid dexamethasone (Dex; 1-50 microg per mouse, i.v., corresponding approximately to 0.03-1.5 mg kg-1, given i.v. 2 h prior to challenge with IL-8) potently inhibited neutrophil infiltration with an approximate ED50 of 5 microg per mouse (~ 0.3 mg kg-1 , i.v.). Passive immunisation of mice with a polyclonal sheep serum raised against the steroid-inducible anti-inflammatory protein lipocortin 1 (LCl)abolished the inhibitory action of Dex whereas a control serum was without effect.6. Local administration of Dex at a dose which was ineffective when given systemically (1 microg) also reduced neutrophil migration induced by IL-8, either alone or in combination with histamine. This local inhibition (~50%), also seen with hydrocortisone (30 microg), was prevented by the concomitant administration of the steroid antagonist RU38486 (10 microg) indicating the involvement of glucocorticoid receptor in the response.7. These findings characterize further the mechanisms underlying PMN recruitment induced by IL-8 in vivo, and point to a role for histamine. The anti-inflammatory action of the glucocorticoids, as in some other models, appears to be LCl-dependent when these drugs are given systemically and LCl independent when the steroids are given locally.