Leptin is known to regulate body weight, energy balance, and reproduction. Therefore, investigation of its physiology is of obvious interest in bulimia nervosa (BN), an eating disorder characterized by body weight-related psychopathology, acute changes in the energy balance, and reproductive alterations. To date, the few studies that have assessed leptin production in BN have had several limitations, including the measurement of blood leptin levels in treated patients and the lack of normal weight healthy controls, so that the information they provide is not conclusive. As the investigation of leptin dynamics is likely to be more informative, we decided to assess leptin response to acute fasting and refeeding in both untreated patients with BN and healthy controls. Twelve women meeting the diagnostic criteria for BN of the Diagnostic and Statistical Manual of Mental Disorders, and 10 healthy women of the same age range participated in a 3-day study. At 1800 h on day 1, they received a meal of 1088 Cal, with 53% carbohydrates, 17% protein, and 30% fat. Then, they fasted until 1800 h on day 2, when they received the same meal. On day 3, they received a standard hospital diet of 2600 Cal, divided into 3 meals, with the same percentages of nutrients as described above. Blood samples were collected at different time points for plasma leptin, glucose, and insulin measurements. In bulimic patients, plasma leptin values were significantly lower than in healthy women (P < 0.0001) and were positively related to body weight, expressed as body mass index (r = 0.86; P < 0.0001). The leptin response to the fasting/refeeding paradigm significantly differed between patients and controls (time x group interaction, P < 0.0001). In fact, in healthy subjects, acute fasting induced a 58% decline in the plasma leptin concentration, whereas such a decrease was only 7% in bulimic women (P < 0.001). After acute refeeding, plasma leptin increased in both groups, although in the patients it did not reach the absolute values observed in normal controls. No significant difference was observed between bulimics and controls in plasma insulin response to the fasting/refeeding paradigm, whereas an abnormal increase in blood glucose levels was observed in the patients after the first meal following acute fasting. We conclude that in untreated women with BN, leptin, despite its very low plasma values, still holds its function as a sensor of body weight changes, but loses its role of signaling acute changes in energy balance.