To evaluate ultrasonography and anthropometry for the measurement of regional body fat in HIV infected patients. In a cross-sectional study, 61 patients receiving antiretroviral therapy underwent ultrasonography and anthropometry for measuring body fat at abdominal, peripheral and facial levels. Reproducibility and accuracy of the measurements at the different compartments were determined using quantitative computed tomography (CT) and clinical evaluation of lipodystrophy as reference standard. Intraabdominal and subcutaneous abdominal fat assessed by ultrasonography correlated with visceral and subcutaneous abdominal fat quantified by CT (r=0.74, P<0.001, and r=0.84, P<0.001, respectively). Ultrasound-determined subcutaneous fat at mid-thigh level correlated with adipose tissue area measured by CT (r = 0.84, P<0.001). Waist-to-hip ratio (r=0.69, P<0.001), suprailiac skinfold (r=0.78, P<0.001) and mid-thigh skinfold thickness (r=0.83, P<0.001) were also significantly correlated with visceral abdominal fat, subcutaneous abdominal fat, and with subcutaneous leg fat quantified by CT, respectively. Poorer correlations were found between ultrasonographic and anthropometric assessments of facial fat, and adipose tissue measured by CT (r=0.15, P=0.25, and r=0.58; P<0.001; respectively). Reproducibility was higher for anthropometry than for ultrasonography in most body regions. The highest variability was observed for ultrasonographic assessment of facial fat (median inter-observer coefficient of variation, 32.10%). Using the clinical diagnosis of lipodystrophy as reference, the best accuracy was observed for ultrasound-determined intraabdominal fat, waist-to-hip ratio and subcutaneous crural fat measured by ultrasonography. Ultrasonography and anthropometry are fairly accurate and reproducible methods for the evaluation of intraabdominal and peripheral fat. Its role for assessing facial fat seems to be more limited.