PurposeResearch on the relationship between sleep duration and obesity defined using multiple anthropometric and bioelectrical indices in women remains scarce. We aimed to explore the association between sleep duration and body mass index (BMI), waist-hip ratio (WHR), body fat percentage (PBF) and visceral fat area (VFA) among females. MethodsWe recruited women for medical examination using multistage cluster sampling. Sleep was assessed using Pittsburgh Sleep Quality Index (PSQI) and sleep duration was categorized into short (<7 h), optimal (7 <9 h) and long sleep (>= 9 h). Weight and height were measured using a calibrated stadiometer. Waist circumference was manually measured. PBF, and VFA were estimated by bioelectrical impedance analysis. Data on sociodemographic characteristics and lifestyle factors were also collected and included in the logistic regression models to explore the independent association between sleep duration and obesity defined by different indices. ResultsA total of 7,763 women with a mean age of 42.6 +/- 13.5 years were included. The percentage of women reporting short and long sleep was 10.3 and 13.4% respectively. The mean BMI, WHR, PBF and VFA were 23.07 +/- 3.30 kg/m(2), 0.78 +/- 0.06, 32.23 +/- 6.08% and 91.64 +/- 35.97cm(2), respectively. Short sleep was independently associated with 35% (95% CI: 1.05-1.75) increased odds of general obesity (BMI >= 28 kg/cm(2)), and long sleep was associated with 18% (95% CI: 1.01-1.37) increased odds of visceral obesity (VFA > 100 cm(2)). No association was observed between sleep deprivation or excessive sleep and high WHR or high PBF. ConclusionIn women, short sleep was associated with an increased odds of general obesity, whereas long sleep was associated with an increased odds of visceral obesity. Longitudinal observations are needed to confirm this cross-sectional relationship.