Adiponectin is an adipose-derived protein with beneficial metabolic effects. Low adiponectin is associated with obesity and related diseases. Significant weight loss increases adiponectin, reducing disease risk. This study compared the effects of two weight-loss diets with different macronutrient compositions on adiponectin. Eighty-one obese women in two cohorts were randomized to a low-fat (LF) or a low-carbohydrate (LC) diet. All subjects underwent equivalent weight-loss intervention, with weight and other measures assessed at baseline and after 6 (cohort I) or 4 (cohort II) months. Body fat was measured by dual energy X-ray absorptiometry. Adiponectin was measured by radioimmunoassay. Diet intake was assessed using 24-h recalls and 3-day diet records. Data were analyzed via t-tests and repeated-measures factorial ANOVA using time, diet, and replicate (cohort I vs. cohort II) as factors. Age, weight, body fat, BMI, adiponectin, and diet were similar at baseline. Following intervention, macronutrient composition of the diet was vastly different between the groups, reflecting the assigned diet. Both groups lost weight and body fat (P < 0.001), with effect in LC dieters greater than LF dieters (-9.1 kg vs. -4.97 kg weight, P < 0.05 and -5.45 kg vs. -2.62 kg fat, P < 0.001). Adiponectin increased in the LC (+1.92 mcg/ml, P < 0.01), but not the LF (+0.86 mcg/ml, P = 0.81), group. There was no correlation between weight loss and increase in adiponectin. These results confirm that diet-induced loss of weight and body fat is associated with increased adiponectin concentrations. This effect is evident with weight loss of 10% or more, and may be greater with LC diets.