We investigated the seasonal variations in the chemical composition and in vivo inflammatory activity of urban air particulate samples in four size ranges (PM(10-2.5), PM(2.5-1), PM(1-0.2), and PM(0.2)). The samples were collected in Helsinki using a high-volume cascade impactor (HVCI). Healthy C57BL/6J mice were intratracheally instilled with a single dose (10 mg/kg) of the particulate samples. The lungs were lavaged and the bronchoalveolar lavage fluid (BALF) was assayed for indicators of inflammation and tissue damage: cytokines (tumor necrosis factor [TNF]-alpha, interleukin [IL]-6, and keratinocyte-derived chemokine [KC]) at 4 h, and total cell number and total protein concentration at 12 h. The PM(10-2.5) and PM(2.5-1) samples had much higher inflammatory potency than the PM(1-0.2) and PM(0.2) samples. The relative inflammatory activities of the autumn samples were the highest on an equal mass basis, but when estimated for the particulate mass per cubic meter of air, the springtime samples had the highest inflammatory potential. Resuspended soil material and other non-exhaust particulate material from traffic were associated with a high inflammatory activity of the PM(10-2.5) and PM(2.5-1) samples. Secondary inorganic ions in the PM(1-0.2) and PM(0.2) samples had inconsistent negative or positive correlations with the inflammatory activity. There were no systematic seasonal variations in the tracers of incomplete combustion and atmospherically oxidized organics in the PM(1-0.2) and PM(0.2) samples, which probably explains their low correlations with the inflammatory activity. In conclusion, in a relatively clean Nordic city, the resuspension of road dust and other non-exhaust particulate material from traffic were the major sources of inflammatory activity of urban air inhalable particles.