To apply the approach of forward genetics (e.g., gene identification with mutagenesis and screening, followed by positional cloning) to the mouse, it is necessary to have available screening tests that can be applied rapidly to individual mice and that give a reliable assessment of visual function. This paper reviews the strengths and limitations of two anatomical tests related to visual function, fundus examination and retinal histological examination. Two tests that do not depend on behavior of a conscious animal are reviewed: the electroretinogram and the visual evoked potentials of the cortex. Eight behavioral tests are also summarized: maze-based tests, cued fear conditioning, tests based on conditioned suppression, visual placing, optokinetic nystagmus, pupillary reflex, and light-induced shifts in circadian phase. It is recommended that retinal histology, the electroretinogram, and visual-evoked potentials be used at the present time for screening because they assess the function and structure of the visual system rapidly and reliably. In fact, the electroretinogram (or visually evoked potentials) can be recorded from several animals simultaneously in response to the same stimulus. It is also recommended that efforts be made to develop more appropriate, automated, behavioral tests of visual perception than are now available, particularly tests that rely solely on rewarding visually evoked behavior. Two other promising behavioral tests are cued fear conditioning and variants of maze tests.