A model for leaf growth, acclimation and senescence is described. The approach is mechanistic and the model, which is carbon-based, has eight state variables: substrate carbon and machinery carbon; primary, secondary and cross-linked cell wall; and two which correspond to parameters of the photosynthesis light response curve--namely, the initial slope and light-saturated photosynthetic rate. The model is used to simulate responses to light and temperature, and predicts quantities such as final leaf area, the time at which the leaf becomes an exporter, senescence time, and the two photosynthetic parameters. The results are in good qualitative agreement with observational responses that have been reported. The model also simulates the effects of changing the light environment of a mature leaf: for instance, it predicts that increasing the light level can increase leaf longevity and vice versa, although, in general, leaves grown continuously in high light have a shorter life than leaves grown continuously in low light.