Routine liposuction has very low perioperative complication rates and is thus considered to be innocuous. Some authors have even proposed that large-volume liposuction could be therapeutic. However, because subcutaneous adipose tissue has nutritional and thermodynamic metabolic functions proportional to the absolute amount and the distribution of fat, it is possible that removal of subcutaneous adipose tissue might be detrimental. We measured the amount of fat removed by large-volume (>1000 cc) liposuction and expressed the results in terms of absolute and relative changes in total body fat and in visceral adipose tissue (nonsubcutaneous adipose tissue) in 63 normal weight to mildly obese women (n = 51) and men (n = 12). Aspiration of 1.5 +/- 0.7 kg (mean +/- SD) of lipid in women removed 9.2 +/- 3.2 percent of body fat or 10.5 percent of subcutaneous adipose tissue corresponding to a 12-percent increase in the ratio of visceral to subcutaneous adipose tissue. One third of the women (n = 17) had a mean increase of 16 percent (range 13 to 21 percent) in the proportion of visceral fat. In the 12 men, aspiration of 1.7 +/- 0.6 kg of lipid removed 9.8 +/- 2.9 percent of body fat or 12.7 +/- 3.6 percent of subcutaneous adipose tissue, resulting in a 14-percent increase in the ratio of visceral to subcutaneous fat. The correlation between aspirate and body mass index was 0.57 (p < 0.001). Although large-volume subcutaneous liposuction removed relatively little body fat, it led to significant increases in the proportion of visceral adipose tissue. Because the proportion of visceral adipose tissue is a risk factor for metabolic complications of obesity, the metabolic effects of large-volume liposuction need to be evaluated.