The accurate assessment of respiration by woody vegetation, still a challenge in plant produc- tivity models, is generally a problem of correctly scaling-up the process from organs to the whole plant. We used a large (41.6 m 3 ), canopy chamber to enclose mature olive trees and to measure aboveground respiration (Rag) under natural environmental condi- tions in an irrigated olive orchard in Cordoba (Spain). The 3-year study assessed nocturnal and seasonal Rag patterns in terms of temperature (T), plant dry matter composition, and phenology. The relative contribu- tions of maintenance and growth respiration to Rag were determined empirically via an independent experiment. Although short-term variations in Rag rates were explained mainly by T variations, over seasonal time-scales this relationship was modu- lated by the vegetative composition of the olive trees and the contribution of growth respiration to Rag when the plants, in different seasons, allocated most of the new assimilates to actively growing shoots, flowers or fruits. Leaf mass and fruit load were the main determinants of Rag ,w hich was weakly affected by differences in woody biomass since woody tissue respiration accounted for just 15 % of Rag. Respiration in olive trees during fruit setting periods is composed of approximately 30 % growth and 70 % maintenance. This study provides an independent evaluation of how, and to what degree, seasonally varying plant organ composition determines total respiration. Improved modelling of ecosystem respiration can be achieved by account- ing for plant biological patterns characterising energy-requiring growth and maintenance pro- cesses, since biochemical kinetics alone cannot explain the observed seasonal variability.