Relatively high-protein diets are effective for body weight loss, and subsequent weight maintenance, yet it remains to be shown whether these diets would prevent a positive energy balance. Therefore, high-protein diet studies at a constant body weight are necessary. The objective was to determine fullness, energy expenditure, and macronutrient balances on a high-protein low-carbohydrate (HPLC) diet compared with a high-carbohydrate low-protein (HCLP) diet at a constant body weight, and to assess whether effects are transient or sustained after 12 weeks.A randomized parallel study was performed in 14 men and 18 women [mean ± SD age: 24 ± 5 y; BMI (in kg/m(2)): 22.8 ± 2.0] on diets containing 30/35/35 (HPLC) or 5/60/35 (HCLP) % of energy from protein/carbohydrate/fat.Significant interactions between dietary intervention and time on total energy expenditure (TEE) (P = 0.013), sleeping metabolic rate (SMR) (P = 0.040), and diet-induced thermogenesis (DIT) (P = 0.027) appeared from baseline to wk 12. TEE was maintained in the HPLC diet group, while it significantly decreased throughout the intervention period in the HCLP diet group (wk 1: P = 0.002; wk 12: P = 0.001). Energy balance was maintained in the HPLC diet group, and became positive in the HCLP diet group at wk 12 (P = 0.008). Protein balance varied directly according to the amount of protein in the diet, and diverged significantly between the diets (P = 0.001). Fullness ratings were significantly higher in the HPLC vs. the HCLP diet group at wk 1 (P = 0.034), but not at wk 12.Maintenance of energy expenditure on HPLC vs. HCLP diets at a constant body weight may prevent development of a positive energy balance, despite transiently higher fullness. The study was registered on clinicaltrials.gov with Identifier: NCT01551238.