The purpose of the study was to determine the agreement of bioelectric impedance (BIA) with dual energy X-ray absorptiometry (DEXA) among Singapore Chinese adults. Hundred subjects [63 healthy males (age: 25.2 +/- 0.5 yrs; weight: 69.7 +/- 1.9 kg; BMI: 22.8 +/- 0.4 kg m(-2)) and 37 females (age: 22.9 +/- 0.3 yrs; weight: 51.8 +/- 0.8 kg; BMI: 20.2 +/- 0.4 kg m(-2) )] were selected and body fat percentage (%BF) was measured by DEXA and BIA. Paired t test and coefficient of correlation statistics were utilized to compare the relationship between %BFBIA and %BFDEXA. Bland and Altman plot was employed to investigate the agreement of %BFBIA with %BFDEXA. The limits of agreement between different methods were defined as mean (M; bias) +/- 1.96SD of the difference between the methods (95% confidence interval; CI). %BFBIA, when compared to %BFDEXA, revealed non-significant underestimation of %BF in females (24.1% <24.4%, p>0.05) and significant overestimation in males (15.5%< 17.7%, p<0.01). There was good absolute agreement between %BFBIA and %BFDEXA among the whole cohort (1.3 +/- 6.9%) as well as among both genders (male: 2.2 +/- 6.7%, female: -0.3 +/- 6.1%) due to small mean differences between both methods. However, wider limits of agreement were revealed for %BFBIA among whole cohort and as well as on gender basis. The results indicate a good agreement between BIA and DEXA in measuring %BF among Singapore Chinese adults, but may not be a suitable method of measuring %BF for clinical purposes among this population due to wider limits of agreement. (C) 2010 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved.