1. Liposomes with conventional and long-circulation times were employed as carriers for the methotrexate derivative MTX-gamma-DMPE (MTX-EPC and MTX-PEG respectively), their mechanism of action was investigated in vitro and in vivo and their therapeutic efficacy assessed using the rat collagen-induced arthritis (CIA) model. 2. At non-toxic dose, both MTX-EPC and MTX-PEG inhibited the lipopolysaccharide (LPS) induced release of IL-1beta from activated rat peritoneal macrophages (rPMPhi) in a dose and time dependent manner. Free methotrexate (MTX) was not active in this respect. After a single intravenous injection (i.v.), and at equivalent doses, both free MTX (500 microg) and MTX-EPC inhibited the LPS induced rise in plasma IL-1beta levels observed in MTX-PEG and saline treated rats. 3. When used to treat established CIA, MTX-EPC resulted in significantly lower clinical score (CS) (1.0+/-0.42 (P<0.001)) and hind paw diameter (HPD) (6.5+/-0.34 mm (P<0.001)) measurements than controls (3.0+/-0.26; 7.33+/-0.41 mm), after only two i.v. doses, and remained significantly lower for the entire experimental period. By day 24 both CS (2+/-0.61 (P<0.001)) and HPD (6.97+/-0.25 mm (P<0.002)) measurements had also become significantly lower in MTX-PEG treated rats than in saline treated controls (3.62+/-0.17, 7. 92+/-0.38 mm) and remained lower until day 30. Joint inflammation in MTX treated rats was completely ameliorated by day 20 but the health and well being of the animals was compromised and the experiment terminated at this time-point. 4. Our results clearly demonstrate that both MTX-EPC and MTX-PEG liposomes have potential for development into therapeutic modalities for the treatment of inflammatory joint disease in man.