Objective:To determine whether the increased plasma leptin levels reported in hemodialyzed patients is a feature of end-stage renal disease or an artifact of hemodialysis, we studied plasma levels in patients treated exclusively by continuous ambulatory peritoneal dialysis (CAPD). Design: Prospective comparison of end points in CAPD patients and matched healthy subjects. Setting: Tertiary care institutional dialysis center. Participants: Fifty-six healthy subjects, age 50.8 ± 2.3 years, body mass index (BMI) 27.7 ± 1.3 kg/m2, recruited through public announcement, and 36 patients with end-stage renal disease, age 51.0 ± 2.4 yr, BMI 28.2 ± 1.3 kg/m2, enrolled in a CAPD treatment program. Intervention: Four exchanges of CAPD per day, using 2.0, 2.5, or 3.0 L of dialysate over a period of 1-96 months (median 22 mth). Main Outcome Measures: The primary outcome measure was plasma leptin concentration. Secondary measures included plasma glucose, insulin, C-peptide, and cortisol concentrations; and residual renal function and dialysis adequacy (Kt/V). Results: Plasma leptin levels in CAPD patients were 27.1 - 490 ng/mL (women) and 1.3 - 355 ng/mL (men); the levels in healthy subjects were 2.0 - 84.7 ng/mL (women) and 1.8 - 55.4 ng/mL (men). The mean leptin levels were 5-fold higher among CAPD-treated men than control men (49.9 ± 18.4 vs 9.8 ± 2.5 ng/mL, p < 0.001) and 7.5-foId higher among CAPD-treated women than control women (220 ± 28.1 vs 29.3 ± 3.7 ng/mL, p < 0.0001). Female gender and BMI were the strongest predictors of hyperleptinemia in CAPD patients. Conclusion: These results indicate that hyperleptinemia is a feature of terminal renal failure, not an artifact of hemodialysis.