Acclimation of respiration to temperature is not well understood. To determine whether whole plant respiration responses to long-term temperature treatments can be described using the Q(10) concept, the CO2 exchange rate of marigolds (Tagetes patula L. 'Queen Sophia'), grown at 20 degrees C or 30 degrees C, was measured for 62 days. When plants of the same age were compared, plants grown at 20 degrees C consistently had a higher specific respiration (R-spc) than plants grown at 30 degrees C (long-term Q(10) = 0.71-0.97). This was due to a combination of greater dry mass at 30 degrees C and a decrease in R-spc with increasing mass. When plants of the same dry mass were compared, the long-term Q(10) was 1.35-1.55; i.e. R-spc was higher at 30 degrees C than at 20 degrees C. Whole plant respiration could be accurately described by dividing respiration into growth and maintenance components. The maintenance respiration coefficient was higher at 30 degrees C than at 20 degrees C, while the growth respiration coefficient was lower at 30 degrees C, partly because of temperature-dependent changes in plant composition. These results suggest difficulties with interpreting temperature effects on whole plant respiration, because conclusions depend greatly on whether plants of the same age or mass are compared. These difficulties can be minimized by describing whole plant respiration on the basis of growth and maintenance components.