A convenient and reliable method (including the detailed design of the apparatus) for determining simultaneously the individual O2 consumption of a group of rats is described. It has approximately the same reliability as the long accepted erythrocyte count determination. By actual use with intact control, thyroidectomized control, and treated thyroidectomized rats, the estimate of error of a single determination (or therefore confidence limits at the 5% level, within which the average would lie if based on a large number of measurements) is [plus or minus] 18% of the mean.