Our aim was to evaluate incidence and riskfactors of liver involvement in obese Italian childrenas assessed by both ultrasonographic and biochemicalparameters. In seventy-five consecutive obese children (age 9.5 ± 2.9 years, males/females41/34), serum levels of enzymes and ultrasonography ofthe liver were evaluated. Tests were repeated one,three, and six months after starting a moderatehypocaloric diet and an exercise program. Three obese childrenwho were found to have chronic viral hepatitis wereexcluded from the study. Thirty-eight of 72 (53%) obesechildren had an ultrasonographic image of bright liver consistent with liver steatosis. Thelatter was severe in nine children, moderate in 16, andmild in 13. Eighteen obese children (25%) had elevatedtransaminase levels. Bright liver andhypertransaminasemia were not due to any of the most common causesof liver disease. Both were rapidly responsive to lossof weight, confirming that liver involvement wassecondary to obesity and that steatosis orsteatohepatitis rather than fibrosis were involved. Obesityduration not more than three years (odds ratio = 4.77),a higher degree of obesity (odds ratio = 2.09), andhypertransaminasemia (odds ratio = 2.15) appeared asimportant predictive factors of liver involvement atultrasonography. Incidence of liver involvement assessedby means of ultrasonography is significantly higher thanthat revealed by measurement of serum liver enzymes. A short duration of obesity emerged as apotentially new risk factor of liver involvement in thepediatric obese population and needs to be confirmed infuture studies.