Optimal resource allocation theory predicts that the photosynthetic capacity of individual leaves (A_max) in a canopy should scale proportionally to the availability of light in the vicinity of a leaf. However, a growing body of literature appears to contradict this prediction. This discrepancy has generally been interpreted as being the results of physiological constraints and ecological trade-offs that impede plants to reach the predicted "true" optimal A_max distribution. Nevertheless, resource allocation theory does not take into consideration the potentially important role of leaf orientation as a modulator of the light environment of individual leaves.  The objective of our study was to find the distribution of leaf traits that maximized whole plant gross carbon gain when both leaf A_max and leaf orientation were allowed to vary with the light environment. Our approach was to do explicit three-dimensional simulations of trees using the program LIGNUM.  In our simulations, both A_max and the orientation of individual leaves changed as a function of the amount of light that was available at the point of emergence of leaves on a branch. Our results show that a gradual increase in leaf inclination with increasing light played a significant role at maximizing whole plant carbon gain even if leaf A_max did not vary. When leaf A_max was also allowed to change, plant photosynthesis was maximized when A_max varied non-proportionally with light availability. These results show that optimal resource allocation theory should explicitly consider leaf orientation in order to predict the natural distributions of leaf traits in canopies.