Objective To investigate effectiveness of anthropometric parameters and body composition analysis for the screening and prediction of metabolic syndrome and explore the best indicator for predicting metabolic syndrome in the elderly. Method A cross-sectional study of 763 (406 men and 357 women) elderly people who participated in the annual health check-up was conducted. Clinical data of all participants were obtained including anthropometric parameters, body composition, lipid profiles, fasting blood glucose, and high sensitivity C-reactive protein. Receiver operating characteristic (ROC) curve was used to determine the optimal cutoff points for waist circumference, waist-to-hip ratio, waist-to-height ratio, percent body fat and fat mass index in relation to the area under the curve (AUC), sensitivity and specificity in the screening and prediction of metabolic syndrome. Result In total subjects, compared with non-metabolic syndrome group,the ROC curve analysis showed that parameters including waist circumference, waist-to-height ratio, waist-to-hip ratio, percent body fat and fat mass index had a significant potential for predicting metabolic syndrome (P<0.001). It was determined that waist circumference of 87.5 cm and 77.5 cm, waist-to-hip ratio of 0.89 and 0.87, waist-to-height ratio of 0.51 and 0.52, percent body fat of 24.1%and 31.7%and fat mass index of 5.00 kg/m2 and 7.80 kg/m2 were the optimal cutoff points for screening and predicting the presence of metabolic syndrome among men and women with a sensitivity of 81.3%,78.8%,87.5%, 51.3%and 83.8%(in men) and 85.1%,79.8%,71.3%, 70.2%and 80.9%(in women) and a specificity of 57.7%,62.6%,50.0%, 75.5%and 51.8%(in men) and 38.0%,53.2%,55.1%, 50.6%and 52.5% (in women),respectively. The area under the ROC curve (AUC) was 0.728, 0.755, 0.716, 0.671 and 0.725 in men and 0.652, 0.707, 0.658, 0.619 and 0.675 in women,respectively. Waist-to-hip ratio showed the highest AUC in all the parameters in men and women. Conclusion Anthropometric parameters and body composition analysis play important roles in the screening and prediction of metabolic syndrome, and waist-to-hip ratio seems to be the best parameter in the screening and prediction of metabolic syndrome in the elderly.