Objective: In the present study we have measured the concentrations of interleukin-6 (IL-6), soluble IL-6 receptor (sIL-6R), tumor necrosis factor-a (TNF-a), interleukin-1b (IL-1b) and IL-1 receptor antagonist (IL-1Ra) in the serum of patients with Graves' disease (GD). By multivariate analysis, we have evaluated the effect of antithyroid treatment, thyroid function, the presence or absence of active thyroid-associated ophthalmopathy (TAO), the patient's smoking habits and the relation to circulating anti-thyrotropin (TSH) receptor (TRAb) and anti-thyroperoxidase antibodies (TPOAb). Subjects: We studied 84 GD patients, 51 untreated and 33 receiving methimazole (MMI) therapy. Twenty-three (45%) untreated patients and 18 (54%) patients on MMI had active TAO. We also studied 67 normal subjects as controls. Thirty-one GD patients (43%) and 16 controls (36%) were smokers. Results: Serum IL-6 concentrations were signi®cantly higher in both untreated patients (P < 0.001) and treated patients (P < 0.006), when compared with controls. Serum sIL-6R concentrations were signi®cantly affected by treatment (P 0:001). Serum IL-1Ra concentrations were not different in GD patients, whether treated or untreated, compared with controls. Serum IL-6 concentrations were not in¯uenced by thyroid function and there was a signi®cant interaction between treatment and the presence of active TAO (P 0:003). In hyperthyroid patients with active TAO serum, sIL-6R concentrations were signi®cantly higher than in those with inactive TAO (P 0:003). In untreated GD patients there was no signi®cant effect of thyroid function and TAO activity on the serum concentrations of TNF-a and IL-1b. Serum IL-1Ra concentrations were not affected by the presence of TAO. Smoking had no effect on serum IL-6, sIL-6R, TNF-a, IL-1b and IL-1Ra concentrations, even in the presence of an active TAO. Serum concentrations of IL-6, sIL-6R, TNF-a and IL-1b and IL-1Ra were not different in patients with and without TRAb or TPOAb, in relation to either thyroid function, TAO activity or smoking. Conclusions: Our work shows that: (i) the proin¯ammatory cytokine pattern in GD is greatly in¯uenced by antithyroid drug treatment; (ii) the increased circulating IL-6/sIL-6R concentrations observed in patients with active TAO may derive from the activation of humoral reactions in sites other than the thyroid; and, (iii) cigarette smoking has no effect on serum IL-1/IL-1Ra concentrations in TAO.