35 1/2-year-old and 31 1-year-old children were screened by two-flash photorefraction for strabismus and refractive errors with and without cycloplegia. The sensitivity of the method to detect refractive errors was tested with an optical demonstration eye. All the children were examined clinically to compare the sensitivity of the method. Every child co-operated with the photography, but 4 children did not co-operate in the clinical examination and were thus excluded from the refractive material, as were also two cases of esotropias which were found. There were no false positive or negative strabismus cases. The total refractive material consisted of 120 eyes. The method was clearly more sensitive for refractive errors with cycloplegia. Even one hyperopia of +5.25 D(OD) and +6.0 D(OS) was underestimated without cycloplegia. In the material there were no cases of anisometropia of over 1.0 D in spherical equivivalent that would cause a potential risk for amblyopia. A rather good correlation of refractive results existed with the method in cycloplegia. With partial overlapping of emmetropic and moderate hyperopic cases. The screening of children aged 1/2-1 year with two-flash photorefraction is simple to perform. The underestimation of symmetrical hyperopias should be accepted when cycloplegia is not used. Only one successful photograph of each child is necessary for the interpretation.