We investigated whether serum leptin levels are elevated in patients with active rheumatoid arthritis (RA) and whether these levels correlate with disease activity. Fifty RA patients were enrolled in this study, and their disease activity was assessed using the disease activity score 28 (DAS28). The patients were divided into two groups according to this score: a high activity group with DAS28 > 3.2 (n = 26) and a low activity group with DAS28 </= 3.2 (n = 24). Serum leptin levels were determined using a primate antibody radioimmunoassay. RA patients with high disease activity had significantly higher mean serum leptin levels, compared to those with low activity (14.2 +/- 10.9 vs. 7.0 +/- 3.4 ng/ml, P < 0.05). Mean leptin levels adjusted according to BMI were 0.6 +/- 0.5 ng m(2)/ml kg for the high activity group and 0.3 +/- 0.2 ng m(2)/ml kg for the low activity group, respectively, which were also significantly different (P < 0.05). Both serum leptin levels and leptin levels adjusted according to BMI correlated well with the DAS28 (r = 0.363 and 0.368, P < 0.05) and CRP levels (r = 0.433 and 0.472, P < 0.05), respectively. Sixteen of the 26 RA patients with high disease activity at the initial assessment were re-evaluated, at which point their DAS28 had decreased to less than 3.2. Mean follow-up leptin level was significantly lower than mean initial leptin level (7.8 +/- 3.7 vs. 16.1 +/- 12.7 ng/ml, P < 0.05). In conclusion, serum leptin levels were higher in RA patients with high disease activity, correlated well with disease activity, and decreased significantly when disease was well controlled.