Inhalation of residual oil fly ash (ROFA) increases pulmonary morbidity in exposed workers. We examined the role of reactive oxygen species (ROS) in ROFA-induced lung injury. ROFA was collected from a precipitator at Boston Edison Co., Everett, MA, USA. ROFA (ROFA-total) was suspended in saline, incubated for 24 h at 37°C, centrifuged, and separated into its soluble (ROFA-sol.) and insoluble (ROFA-insol.) fractions. Sprague-Dawley rats were intratracheally instilled with saline or ROFA-total or ROFA-sol. or ROFA-insol. (1 mg/100 g body wt.). Lung tissue and bronchoalveolar lavage cells were harvested at 4, 24, and 72 h after instillation. Chemiluminescence (CL) of recovered cells was measured as an index of ROS production, and tissue-lipid-peroxidation was assessed to determine oxidative injury. Significant amounts of Al, Fe, and Ni were present in ROFA-sol., whereas ROFA-insol. contained Fe, V, and Al. Using electron spin resonance (ESR), significantly more hydroxyl radicals were measured in ROFA-sol. as compared to ROFA-insol. None of the ROFA samples had an effect on CL or lipid peroxidation at 4 h. Treatment with ROFA-total and ROFA-insol. caused significant increases in both CL (at 24 h) and lipid peroxidation (at 24 and 72 h) when compared to saline control value. ROFA-sol. significantly reduced CL production at 72 h after treatment and had no effect on lipid peroxidation at any time point. In summary, ROFA, particularly its soluble fraction, generated a metal-dependent hydroxyl radical as measured by a cell-free ESR assay. However, cellular oxidant production and tissue injury were observed mostly with the ROFA-total and ROFA-insol. particulate forms. ROS generated by ROFA-sol. as measured by ESR appear not to play a major role in the lung injury caused after ROFA exposure.