OBJECTIVE: To determine whether the kinetics of protein metabolism would differ with the prolonged use of a casein-soy 96 g protein, 1.7MJ/d very low energy diet (VLED) from those of a tryptophan- and methionine-supplemented hydrolyzed collagen VLED, in obesity. DESIGN: Clinical intervention study of 1 week isoenergetic diet followed by 6 weeks VLED. SUBJECTS: 6(1M,5F) healthy obese subjects (age: 38+/-4y, BMI: 33+/-1kg/m(2), weight: 97+/-7kg). MEASUREMENTS: Whole-body nitrogen (N) flux rate (Q) from N-15 abundance in urinary urea using the oral N-15-glycine method and rates of protein synthesis (S) and breakdown (B) calculated from Q;N balance; resting metabolic rate; metabolic and hormonal responses. RESULTS: Q(per kg LBM) was maintained with both collagen and casein-soy VLED. S and B decreased (P<0.05) at week 4 of both VLEDs with resulting decreases in net protein synthesis. At week 6, S decreased with both VLEDs, but B decreased only with casein-soy, at which time N balance was at equilibrium with casein-soy but returned to negative with collagen. Initial resting metabolic rate correlated with baseline Q and S. It decreased by 16% with the VLEDs; 25% of the decrease may derive from the decline in S. CONCLUSION: A 6 week casein-soy VLED with 46% of amino acids as essential does not provide a substantial advantage compared to hydrolyzed collagen with 16% of amino acids as essential. With prolonged use, it may better maintain N equilibrium by preventing further increments in protein breakdown.