Tradit~onally, field studies of size selective predation by visual-feeding planktivorous fish compare the size distribution of prey in stomachs to that of zooplankton measured in the field However, the size frequency of prey perceived by a fish may differ from that measured by researchers using integrative plankton nets. In this study, a mechanistic spatial foraging model was developed to test the hypothesis that size-selective predation by visual-feeding planktivorous fish can arise simply by random encounters with prey. Our model was based on movement of a single predator and a size structured prey population distributed in 3-dimensional space. The model assumed that a predator's encounters with prey was a function of prey size and predatol-swimming speed. Upon encounter, the predator either selected prey by random choice or the largest apparent size. The size frequency and distance distribution of prey encountered, selected, and captured were estimated using Monte Carlo techniques. The model performance was initially evaluated wlth hypothetical prey slze frequencies, and different light attenuation coefficients, capture efflciencies, visual distances and predator swimming speeds. Simulation results showed that the size frequency of prey encountered randomly by a predator is very different from the size frequency of the a i n b~e n t prey In the environment The s l r c frequency of prey selected by random c h o~c c only differed from that selected by appal-ent size choice when visual distance was greater than reactive distance. Prey distance distribution, which was defined as frequency of prey encountered a t different distances, showed that as prey density increased predators sclected prey at closer distances by the largest apparent size choice than by random choice. We also tested the model specifically for the bay anchovy Anchoa mitchilli. Ambient zooplankton size frequencies found in m~d-Chesapeake Bay were used to predict the size frequencies of zooplankton consumed by fish. Predicted prey size frequency in the diet matched the size frequency of zooplankton found in bay anchovy stomachs. We conclude that prey slze selection by fish can be described mechanistically by differentlal random encounter from a flsh's perspective, and that behavioral choice plays a minor role In prey size selection.