In principle, the respiratory loss associated with the synthesis of plant biomass from a given amount of available photosynthate should be independent of temperature, while the maintenance loss rate should increase geometrically with increasing temperature. When the whole plant carbon balance is considered, the daily carbon gain with a diurnal cycle of temperature should therefore be smaller than with a constant temperature of the same mean, which is the opposite of that commonly assumed.The carbon balances of dense stands of white clover (Trifolium repens L. ‘Regal’) were determined from integrated CO2 exchange rates (CER). The carbon balances with a diurnal cycle of temperature (30 C day/20 C night, 30 C day/10 C night) were shown to be predictable from the balances at constant temperatures of 30 C, 20 C, and 10 C. With a 30 C day, the daily carbon gain (growth rate)increased as night temperature decreased. The gain with a 30/10 C diurnal cycle was less than that with a constant 20 C. These results agree with the above principles.The daily substrate input of carbon from photosynthesis, calculated as the integral of the CER in the light minus the CER in the dark during the daytime, was the same with the 30/10 C diurnal cycle as with a constant 20 C. The CER in the light (the so‐called “net photosynthetic rate”) was much smaller with the diurnal cycle, while the daytime CER in the dark was much more negative. This was attributed to accelerated use of substrate that could not be used during the cool night.Hourly changes in the CER and in the levels of the various pools of substrate were simulated using a dynamic carbon balance model after Thornley. Parameters of the model were varied until the simulated pattern of the CER matched the observed pattern at a constant 30 C, then the rate of use of substrate and the rate of degradation of biomass were reduced by a factor of 2 for each 10 C drop in temperature, day or night.There was good agreement between simulated and observed CER for all combinations of temperature tested. Minor differences were attributed to deviations from the assumed steady state of substrate production and use, and to accelerated senescence of leaves at 30 C. An additional limitation is that developmental effects were excluded. Nevertheless, it is clear that the daily carbon balance with a diurnal cycle can be expected to differ only slightly from the balance with a constant temperature having the same mean value.