Female Pekin ducks selected for high (lean) or low (fat) breast muscle:total breast thickness ratio (MT:TOT), measured with ultrasound scanning, were used in metabolic studies to determine the metabolic differences and assess the effects of feed withdrawal and refeeding on plasma parameters. Lean and fat ducks had similar body weight changes throughout the studies. Lean ducks had significantly lower fat and higher CP and ash carcass contents. They also had higher (P < 0.05) plasma glucose after feed withdrawal but less (P < 0.05) plasma glucose after feeding. Fat ducks had similar plasma triglycerides (TG) after feed withdrawal but higher (P < 0.05) TG concentration after feeding and higher (P < 0.05) plasma uric acid than the lean females after feeding. Feed deprivation resulted in higher (P < 0.05) plasma cholesterol than did feeding. Lean ducks fed the test diet retained more (P < 0.05) N than the fat birds. Lean and fat Pekin ducks had similar (P > 0.05) AME, AMEn, TME, and TMEn. Lean ducks lost less (P < 0.05) N than fat birds. Plasma cholesterol was not significantly correlated (P = 0.06) with nitrogen retention (NR); NR was negatively correlated (P < 0.05) with plasma TG, and TG was positively correlated (P < 0.05) with total energy excreted. There was a positive correlation (P < 0.05) between MT:TOT and NR. The MT:TOT correlated (P < 0.05) negatively with carcass fat, and positively with carcass CP and plasma insulin-like growth factor I (IGF-I) concentration. The results we obtained demonstrate the usefulness of ultrasound scanning in live bird selection for leanness and greater breast muscle thickness and are related to metabolic differences between lean and fat female Pekin ducks selected for high or low MT:TOT ratios.