We questioned whether the amplitudes of the circadian pattern of body temperature ( T b), oxygen consumption ( V˙ O 2 ) and heart rate (HR) changed systematically among species of different body weight ( W). Because bodies of large mass have a greater heat capacitance than those of smaller mass, if the relative amplitude (i.e., amplitude/mean value) of metabolic rate was constant, one would expect the T b oscillation to decrease with the increase in the species W. We compiled data of T b, V˙ O 2 and HR from a literature survey of over 200 studies that investigated the circadian pattern of these parameters. Monotremata, Marsupials and Chiroptera, were excluded because of their characteristically low metabolic rate and T b. The peak-trough ratios of V˙ O 2 (42 species) and HR (35 species) averaged, respectively, 1.57±0.08, and 1.35±0.07, and were independent of W. The daily high values of T b did not change, while the daily low T b values slightly increased, with the species W; hence, the high-low T b difference (57 species) decreased with W (3.3 °C· W −0.13). However, the decrease in T b amplitude with W was much less than expected from physical principles, and the high-low T b ratio remained significantly above unity even in the largest mammals. Thus, it appears that in mammals, despite the huge differences in physical characteristics, the amplitude of the circadian pattern is a fixed (for V˙ O 2 and HR), or almost fixed (for T b), fraction of the 24-h mean value. Presumably, the amplitudes of the oscillations are controlled parameters of physiological significance.