In this study, we sought to determine the relationship between serum levels of leptin and adiponectin (Acrp30) in patients with HIV-associated lipodystrophy (HIV-LD). Three groups of subjects were studied; HIV-positive subjects with lipodystrophy (HIV-LD; n = 22), HIV-positive subjects without lipodystrophy (HIV; n = 17), and ethnicity- and body mass index-matched healthy control subjects (n = 20). Although total body fat from dual energy x-ray absorptiometry was similar in all three groups, the HIV-LD group had a significantly lower mean proportion of body fat in the limbs +/- SEM (37.2% +/- 2.2%) than either controls (49.8% +/- 1.5%) or HIV subjects (45.7% +/- 2.0%). The HIV-LD group also had the lowest mean insulin sensitivity +/- SEM (5.11 +/- 0.59 mg of glucose/[kg of lean body mass. min] vs. 10.2 +/- 0.72 mg of glucose/[kg of lean body mass. min] in controls and 8.64 +/- 0.69 mg of glucose/[kg of lean body mass. min] in the HIV group). Leptin levels were similar in all three groups and were significantly correlated to total body fat (r = 0.86; p <.001), but these levels did not correlate with either insulin sensitivity or limb fat. Mean Acrp30 levels +/- SEM were lowest in the HIV-LD group (5.43 +/- 0.44 microg/mL vs. 11.2 +/- 1.4 microg/mL in the HIV group and 14.9 +/- 1.8 microg/mL in control subjects). Further, Acrp30 levels were positively correlated with insulin sensitivity (r = 0.610; p <.001) and limb fat (r = 0.483; p <.001). However, the correlation between limb fat and insulin sensitivity disappeared when Acrp30 level and other potential mediators were removed from the association, suggesting that a deficiency in Acrp30 in subjects with HIV-LD may be part of the mechanism for the reduced insulin sensitivity.