Light absorption and use efficiency (LAUE mol mol(-1), daily gross photosynthesis per daily incident light) of each leaf depends on several factors, including the degree of light saturation. It is often discussed that upper canopy leaves exposed to direct sunlight are fully light-saturated. However, we found that upper leaves of three temperate species, a heliophytic perennial herb Helianthus tuberosus, a pioneer tree Alnus japonica, and a late-successional tree Fagus crenata, were not fully light-saturated even under full sunlight. Geometrical analysis of the photosynthetic light response curves revealed that all the curves of the leaves from different canopy positions, as well as from the different species, can be considered as different parts of a single non-rectangular hyperbola. The analysis consistently explained how those leaves were not fully light-saturated. Light use optimization models, called big leaf models, predicted that the degree of light saturation and LAUE are both independent of light environment. From these, we hypothesized that the upper leaves should not be fully light-saturated even under direct sunlight, but instead should share the light limitation with the shaded lower-canopy leaves, so as to utilize strong sunlight efficiently. Supporting this prediction, within a canopy of H. tuberosus, both the degree of light saturation and LAUE were independent of light environment within a canopy, resulting in proportionality between the daily photosynthesis and the daily incident light among the leaves.