This study assesses the diagnostic accuracy of Body Mass Index (BMI) in identifying obesity in a large, multiethnic sample of US adults. Using bioelectrical impedance analysis (BIA) to estimate body fat percentage (BF%), the researchers found that while BMI has a good correlation with BF%, it has limited diagnostic performance, particularly for individuals with intermediate BMI ranges (25-30 kg/m²). A BMI cut-off of ≥ 30 kg/m² had high specificity but poor sensitivity, missing more than half of those with BF%-defined obesity. The diagnostic performance of BMI diminished with age, and it correlated better with lean mass than with BF%. These findings suggest that BMI may underestimate the prevalence of obesity and that direct measures of body fatness and body fat distribution may be more accurate in intermediate BMI ranges. The study highlights the need for alternative methods to better diagnose obesity and predict obesity-related health risks.This study assesses the diagnostic accuracy of Body Mass Index (BMI) in identifying obesity in a large, multiethnic sample of US adults. Using bioelectrical impedance analysis (BIA) to estimate body fat percentage (BF%), the researchers found that while BMI has a good correlation with BF%, it has limited diagnostic performance, particularly for individuals with intermediate BMI ranges (25-30 kg/m²). A BMI cut-off of ≥ 30 kg/m² had high specificity but poor sensitivity, missing more than half of those with BF%-defined obesity. The diagnostic performance of BMI diminished with age, and it correlated better with lean mass than with BF%. These findings suggest that BMI may underestimate the prevalence of obesity and that direct measures of body fatness and body fat distribution may be more accurate in intermediate BMI ranges. The study highlights the need for alternative methods to better diagnose obesity and predict obesity-related health risks.