A reasonable amount of information concerning early refraction and accommodation has been accumulated. The measurements of refraction appear to be reliable and concurrently valid, although more attention should be devoted to both of those issues. There is also some evidence that the measurements can be developmentally significant since adults with histories of early spherical or cylindrical errors exhibit nonreversible visual deficits. The retrospective findings [11, 43] that evidenced this point do not necessarily imply that large refractive errors during infancy invariably yield visual deficits later on. To determine the significance of early errors, their predictive validity must be better estimated. Some evidence indicates that neonatal refractions predict refractions later in infancy. But the actual magnitude of the relationship between early and school aged or adult refractions is uncertain. Longitudinal studies of refraction from birth to at least the elementary school years would yield appropriate data. It is hoped that with the increasing interest in early visual function and the development of accurate, yet rapid techniques of measurement (such as photorefraction), such information will soon be forthcoming. It will be of special interest to unveil the mechanisms that control the development of refraction, in particular, to reveal the developmental relationship between refraction and accommodation.