The application of detailed models of canopy photosynthesis rely on the estimation of attenuation of light in the canopy. This attenuation is readily estimated with the Lambert-Beer law when the canopy is homogeneous. In reality, forest canopies are far from homogeneous, and this has led to the use of detailed light extinction models that account for grouping of foliage between and within trees. Because such models require detailed parameterization and fine resolution inputs, they are impractical in larger-scale applications. Thus, there is interest in simplified models that can be readily parameterized. We developed two equations that can be used to estimate mean annual light interception by single unshaded trees and by stands of Poisson distributed trees. Interception by single trees is a function of crown surface area, the ratio of leaf area to crown surface area, the extinction coefficient in a homogeneous canopy--which can be determined separately--and one empirical parameter that depends on the mean solar angle. The summary model was tested against a detailed model of interception, and showed good agreement, although with slight bias. The results showed that crown surface area is a good summary variable for crown size and shape, because errors are independent of crown shape (ellipsoids, cones and height:width ratios). We also tested whether canopy photosynthesis is proportional to light interception across canopies differing in structure and leaf area index, and found that light-use efficiency is influenced by canopy structure. The model is useful in larger-scale applications because it can be parameterized with available data without the need for additional empirical parameters. It can also be used to study the effect of stand structure on mean annual light interception and productivity.