1 Anti-in¯ammatory e ects of cyclic AMP elevating agents were examined in a mouse model of lipopolysaccharide (LPS)-induced microvascular permeability change. 2 Vascular permeability on the back skin was measured by the local accumulation of Pontamine sky blue (PSB) after subcutaneous injection of LPS (400 mg site 71 ) from Salmonella typhimurium. 3 Dye leakage in the skin was signi®cantly increased 2 h after injection of LPS. This LPS-induced dye leakage was suppressed by phosphodiesterase inhibitors, including pentoxifylline (160 mg kg 71 ), milrinone (5 ± 10 mg kg 71 ), rolipram (0.5 ± 10 mg kg 71 ) and zaprinast (5 ± 10 mg kg 71 ). The dye leakage was also inhibited by b-adrenoceptor agonists, including isoproterenol (0.5 ± 5 mg kg 71 ) and salbutamol (0.05 ± 5 mg kg 71 ), an adenylate cyclase activator, forskolin (5 mg kg 71 ), and a cell permeable cyclic AMP analogue, 8-bromo-cyclic AMP (8-Br-cAMP, 10 mg kg 71 ). 4 LPS caused a transient increase in serum TNF-a level peaking at 1 h after the injection. This increase in serum TNF-a was completely blocked by a pretreatment with pentoxifylline (160 mg kg 71 ), milrinone (5 mg kg 71 ), rolipram (1 mg kg 71 ), zaprinast (10 mg kg 71 ), salbutamol (0.5 mg kg 71 ), forskolin (1 mg kg 71 ) and 8-Br-cAMP (10 mg kg 71 ). 5 LPS caused an increase in serum IL-1a level peaking at 3 h after injection. This increase in serum IL-1a was not signi®cantly suppressed by the cyclic AMP elevating agents. 6 Our study suggests that cyclic AMP elevating agents attenuate LPS-induced microvascular permeability change by suppressing TNF-a up regulation.