Voluntary intake of protein, fat and carbohydrate (CHO) was modified by feeding young rats either a control purified diet [% metabolizable energy (ME): protein 21, fat 7, CHO 72], a control diet plus sucrose solution (20%) to drink (final intakes 17, 6 and 77% ME as protein, fat and CHO, respectively) or a low protein diet substituted with either CHO (8, 7 and 85% ME as protein, fat and CHO, respectively) or fat (8, 20 and 72% ME as protein, fat and CHO, respectively). Total ME intakes corrected for body size were similar for all rats, but body weight, energy gain and net energetic efficiency were lower in both low protein-fed groups than in the control group. The acute thermogenic response (% rise in oxygen consumption) to a standard balanced-nutrient meal was higher (12%) in sucrose-supplemented and in low protein groups (15-16%) than in control rats (8%). Brown adipose tissue protein content and thermogenic capacity (assessed from purine nucleotide binding to isolated mitochondria) were greater than control values in sucrose-fed and protein-deficient animals, and the greatest levels of activity were seen in low protein-fed rats with a high fat intake. The results demonstrate that the changes in energy balance, thermogenesis and brown adipose tissue activity that result from protein deficiency cannot be ascribed to changes in the level of energy intake or to a specific increase in the amount or proportion of either CHO or fat. They suggest that the protein-to-energy ratio must be the primary influence on thermogenesis and brown fat activity in these animals.