This study aims to investigate anthropometric measures and their effectiveness as screening method for diagnosing obstructive sleep apnea (OSA) in each gender. We also evaluated which measures were associated with OSA in the adult population of a large metropolitan city, Sao Paulo, Brazil. 552 women and 450 men were submitted to polysomnography (PSG), and the anthropometric measurements as body mass index (BMI), waist-to-height ratio, neck and waist circumference were collected. The measurements were then compared with the OSA classification established by the PSG. In women, waist circumference and waist-to-height ratio were found to be the best predictor, while in men, the factors with great potential for identification varied according to severity of the disease, highlighting waist-to-height ratio, neck circumference and BMI had strongest association. The accuracy of the classification in relation to mild-to-severe OSA based on cut-off values of 92.5cm for waist circumference was greater than 72.9% in men, and 78.9% in women based on cut off values of 95cm. Regarding severe OSA, cut-off values of 116.1cm were greater than 91.3% accurate in the male population, and 95.1% in the female population with a cut-off value of 126.5cm. The study found waist circumference and waist-to-height ratio to be the best measure to assess sleep-disordered breathing in women. Waist-to-height ratio and neck circumferences were the best measures in men with mild OSA, but BMI was more closely associated with severe OSA. The present study identified the anthropometric variables with the highest risk for OSA and their respective cutoff value, according to gender.