1. We have studied the development of the refractive state in young barn owls (Tyto alba pratincola). Strikingly, the eyes had severe refractive errors shortly after lid opening (which occurred around day 14 after hatching; average from 6 owls: 13.83 _ 1.47 days). Refractive errors vanished in the subsequent one or two weeks (Fig. 1, Fig. 2). 2. Refractive errors did not differ by more than 1 diopter (D) in both eyes of an individual (Fig. 2). Thus, non-visual control of eye growth was sufficient to produce non-random refractions. However, visual input was finally required to adjust the optical system to emmetropia. 3. Using in-vivo A-scan ultrasonography of ocular dimensions (Fig. 4A), photokeratometric measurements of corneal radius of curvature (Fig. 4B), and frozen sections of excised eyes (Fig. 3), we developed paraxial schematic eye models which described age-dependent changes in ocular parameters and were applicable through the ages from lid opening to fledging (Table 1). A schematic eye for the adult barn owl (European subspecies: Tyto alba alba) is also provided. Eye sizes in an adult owl of the American (Tyro alba pratincola) and the European subspecies (T. alba alba) were similar despite of different body weights (500 g and 350 g, respectively). 4. The schematic eyes were used to test which ocular parameters might have caused the recovery from refractive errors. However, none of the ocular dimensions measured underwent obvious changes in their growth curves as visual input became available. Apparently, F. Schaeffel ([5~)