Loss of body fat by caloric restriction is accompanied by decreased circulating leptin levels, increased ghrelin levels, and increased appetite. In contrast, dietary fat restriction often decreases adiposity without increasing appetite. Substitution of dietary carbohydrate for fat has been shown to increase the area under the plasma leptin vs. time curve (AUC) over the course of 24 h. This effect, if sustained, could explain the absence of a compensatory increase in appetite on a low fat diet. To clarify the effect of dietary fat restriction on leptin and ghrelin, we measured AUC for these hormones in human subjects after each of the following sequential diets: 2 wk on a weight-maintaining 35% fat (F), 45% carbohydrate (C), 20% protein (P) diet (n = 18); 2 wk on an isocaloric 15% F, 65% C, 20% P diet (n = 18); and 12 wk on an ad libitum 15% F, 65% C, 20% P diet (n = 16). AUC for leptin was similar on the isocaloric 15% F and 35% F diets (555 +/- 57 vs. 580 +/- 56 ng/ml.24 h; P = NS). Body weight decreased from 74.6 +/- 2.4 to 70.8 +/- 2.7 kg on the ad libitum 15% F diet (P < 0.001) without compensatory increases in food consumption or AUC for ghrelin. Proportional amplitude of the 24-h leptin profile was increased after 12 wk on the 15% fat diet. We conclude that weight loss early in the course of dietary fat restriction occurs independently of increased plasma leptin levels, but that a later increase in amplitude of the 24-h leptin signal may contribute to ongoing weight loss. Fat restriction avoids the increase in ghrelin levels caused by dietary energy restriction.