Indirect calorimetric measurements have been used for years to determine energy expenditure (EE), respiratory quotient (RQ) and substrate oxidation rates of the subjects studied. This technique has now been extended to solve the problem of estimating locomotor activity (LA) and utilisation factors (UF). The method assumes (a) the organism rests at least once in a defined period, (b) this rest is long enough to reach basal CO2 and O-2 concentrations in the calorimetric chamber, and (c) data acquisition occurs often enough to capture those minimum values. Connecting and smoothing these minima separates LA from EE and leads to the time course of resting metabolic rate (RMR) + postprandial thermogenesis (ppTh). If the organism reaches a postabsorptive state without experiencing deep hunger, the ppTh is zero and so this value can be interpreted as RMR. Some tests gave the following relative errors (r = s(d)/(x) over bar) showing the reliability of this technique : rho(EE) = 3.2%, rho(RQ) = 1.7%, rho(LA) = 4.6%, rho(RNU) = 4.6%. Under some circumstances it is very useful to know to what extent a given amount of energy can be utilised. We developed a scheme to estimate energy and substrate UF and tested it with a diet comparable to one for prematurely born babies. Rats were placed in the indirect calorimetry unit for 4 days and were given that diet in the order of 450, 600, 750 and 0 kJ kg(-0.75). Energy and substrate balances (BA = E(intake) - E(expenditure)) were calculated for each individual and each energy intake (EI). Individual linear regression analyses BA = f(EI) were calculated, giving estimation factors (r(2)) of 99.71 +/- 0.13% for energy utilisation. The correlation factors of that regression were 80.5 +/- 4.4% with a relative reliability of approximate to 5.3%. They were then interpreted as UF.