We investigated the role of tumour necrosis factor-alpha (TNF) in the evaluation of pleural effusion aetiology. Using a commercially-available ELISA kit, concentrations of TNF were measured in the serum and pleural fluid of patients with malignant effusions (n=19), uncomplicated parapneumonic effusions (n=13), and exudative (n=13) and transudative (n=13) effusions due to congestive heart failure (CHFex and CHFtr, respectively). Serum TNF did not differ significantly between the four groups (p>0.05). In the group with malignancy, pleural fluid TNF was significantly higher than in the other groups (p<0.001), which were not significant different from each other (p>0.05). However, a considerable overlap between all four groups was found. Pleural fluid TNF was significantly higher than serum TNF in the malignant and the uncomplicated parapneumonic groups (p<0.001), and there was a significant positive correlation between serum TNF and pleural fluid TNF in the group with uncomplicated parapneumonic effusion (r=0.7, p<0.005), in the group with CHFex (r=0.54, p<0.01), and in the group with CHFtr (r=0. 8, p<0.005), but not in the group with malignancy. Pleural fluid TNF:serum TNF (TNF ratio) was significantly higher in the malignancy group than in the other groups (p<0.001); no significant difference was found between the other three groups (p>0.05). At an optimal cut-off point of 2.0 for TNF ratio, determined by ROC analysis for discrimination between malignant and non-malignant groups, sensitivity was 84%, specificity 90%, and total accuracy 88% (p<0. 0001). TNF ratio might be helpful in the diagnostic assessment of exudative pleural effusion.