Background: Metabolic syndrome (MetS) and visceral obesity are more prevalent in obstructive sleep apnea syndrome (OSAS). We investigated the association of visceral fat (VF) measures with the components of MetS in OSAS patients with different severity levels, according to World Health Organization (WHO, 1999), National Cholesterol Education Program Adult Treatment Panel III (NCEP-ATP III, 2001), and International Diabetes Federation (IDF, 2005) definitions. Patients and Methods: Study population was grouped according to polysomnography results as non-OSAS [who had apnea-hypopnea index (AHI) <5, n=51], mild OSAS (5 < AHI <15, n=52), moderate OSAS (15 < AHI <30, n=53), and severe OSAS (AHI >30, n=53). VF ratio was measured by abdominal bioimpedance analysis. Waist-to-hip ratio (WHR), homeostasis model assessment of insulin resistance (HOMA-IR), and lipid profiles were assessed in all subjects. Results: The prevelance of MetS in OSAS patients was 30.0%, 35.6%, and 44.4% according to WHO, NCEP-ATP III, and IDF definitions, respectively. MetS was found in 27.5% non-OSAS and 72.8% OSAS according to at least one definition (P=0.012). Within OSAS group, 27.2% subjects had average, 38.0% had slightly excessive, and 34.8% had an excessive VF ratio. The prevelance of MetS was similar in various VF ratios (P>0.05). However HOMA-IR increased progressively with VF ratio after adjusting for age, gender, and body mass index (BMI; P=0.02). AHI increased progressively with BMI (P=0.02), WHR (P=0.03), VF ratio (P=0.01), HOMA-IR (P=0.02), and MetS (P=0.016). Conclusion: Since severity of OSAS, in terms of AHI and insulin resistance, is both associated with VF rather than BMI, VF should be suggested to link OSAS and MetS. The IDF definition is more sensitive in OSAS patients to diagnose MetS, as central obesity and insulin resistance are obligatory components. This would allow clinicians to intervent earlier to adverse metabolic outcomes of OSAS.