Endotoxins are potent pro-inflammatory substances present in several natural environments and in commercial house dust extracts. To investigate the possible effect of chronic endotoxin exposure on asthma, 28 patients with perennial chronic asthma (20 allergic to house dust mite and eight intrinsic asthmatics) were evaluated during a 4-month period (lung function, clinical and immunological criteria). At the same time, two house dust samples were collected from each patient's home to determine total house dust weight (mg/m2), endotoxin concentration and house dust mite antigen content (evaluated indirectly by guanine content with HPLC method). The mean (+/- s.d.) endotoxin concentration, as measured by quantitative Limulus assay was 2.59 (+/- 3.41) ng/mg house dust, ranging from 0.12 to 20 ng/mg. The mean guanine content was 0.13 (+/- 0.16) mg/100 mg house dust. There was no correlation between endotoxin and house dust mite concentrations. Patients were compared according to the low or high grade exposure to dust, endotoxins and guanine. Compared with patients with low grade (less than or equal to 5.6 ng/ml) exposure, subjects exposed to high endotoxin concentrations (greater than 5.6 ng/ml) showed a significant increase in dyspnea (median 2.6 vs 3.3; P less than 0.05) and treatment (median 14 vs 44.3; P less than 0.01) scores, oral corticosteroid (median 0.0 vs 13.5 mg/24 hr; P less than 0.01) and beta 2-mimetics (median four vs eight puffs/day; P less than 0.01) intake, and a significant decrease in FEV1/FVC (median 84.5 vs 67% of predicted value; P less than 0.01).(ABSTRACT TRUNCATED AT 250 WORDS)