While visceral fat measured by dual-energy X-ray absorptiometry (DXA) is accurate in identifying middle-aged people at increased cardiometabolic risk, consistent data for the elderly are still lacking. We aimed to investigate the association between DXA-derived visceral adipose tissue (VAT) and metabolic syndrome (MetS) and to establish optimal cutoffs for VAT to predict MetS in a low-income elderly Brazilian cohort. A total of 449 women and 258 men (≥65 years) from the community were enrolled in this study. Participants underwent clinical and laboratory evaluations, along with body composition analysis by Hologic Discovery A densitometer. VAT was measured in the android region of the DXA scan. MetS was diagnosed using NCEP-ATPIII criteria. Multivariate logistic regression analyzed the relationship between VAT and MetS. Receiver-operating characteristic (ROC) curve analysis evaluated VAT's predictive accuracy for MetS, with optimal cutoffs determined by Youden's test to balance sensitivity and specificity. Mean ages were 76.6 ± 4.7 years for men and 77.1 ± 4.9 years for women. Mean BMIs were 26.5 ± 3.8 kg/m2 for men and 29.0 ± 5.2 kg/m2 for women. One hundred and seventy-five (41.5 %) men and 274 (61 %) women had MetS. After adjustments for confounders, multivariate analysis showed that VAT was independently associated with MetS in both men (OR 1.41, 95%CI 1.15-1.72) and women (OR 1.33, 95%CI 1.16-1.54, per each 100 g increase). Optimal VAT cutoffs to predict MetS were 642.5 g for men (AUC = 0.740) and 600.5 g for women (AUC = 0.729). Subanalysis for non-overweight/non-obese subjects yielded lower VAT cutoffs. Thus, VAT measured by DXA was significantly associated with MetS in older adults, regardless of BMI, emphasizing the critical role of VAT in predicting MetS. Therefore, VAT by DXA holds promise for evaluating MetS risk in the elderly. Further longitudinal studies are needed to investigate VAT's impact on major cardiovascular event incidence in this demographic.