This paper describes the first application of structural modeling to the visual system. Structural modeling, or path analysis, is a mathematical method that allows for the quantification of the functional strengths of anatomical connections between the structures that form a neural system. The objective was to demonstrate how structural modeling can be used to determine the functional interrelationships between brain structures that form the visual system and how these interrelationships change under different conditions. Data were obtained from measures of 2-deoxyglucose uptake in the visual system of rats presented with either patterned light or darkness. The effects of arousing footshock on visual system operations were also investigated. Models based on the anatomical connections and the interregional correlations between metabolic activity data were used to determine path coefficients representing the magnitude of the influence of each directional path. Statistical evaluation of the models revealed that the dominant positive influences on visual system activity in the darkness were the tectocortical subsystem and the descending connections from secondary visual cortex. In the patterned light model, the total influence of the geniculocortical subsystem was higher than in the dark, and the tectocortical pathways showed both a reduction and a shift in the direction of effects. The models also revealed that the effects of footshock-induced arousal on visual system operations depended upon the visual environment and on extra-visual influences. The footshock led to an increase in the interaction of the two main subsystems at the level of connections between primary visual cortex and the lateral posterior nucleus, and a descending negative influence from the secondary visual cortex became dominant. The models are discussed in the context of conventional analyses to show how structural modeling allows for the determination of much more information about the functional interactions within the visual system of subjects under different experimental conditions.