RELIABLE statistics of any type, properly collected, used and interpreted, can cast their shadow into the future. Only lately has there emerged the possibility of utilizing this tool to greater degree for prediction than was formerly the case in the field of blindness. But what are reliable blindness statistics and how do we arrive at them? Over the last few months I have reviewed the worldwide literature on the incidence and prevalence of blindness, and have been amazed at the lack of comparability of data. Quite often the studies fail to even mention the definition of blindness used. Where mentioned, there appears to be no uniformity of agreement among countries concerning the definition of blindness. Legal blindness, economic blindness, industrial blindness-these may mean different things in different countries and sometimes even in different geographic areas in the same country. In Finland a person is blind who cannot find his way in an unknown place for lack of sight. In Iran he is blind when he is deprived of sight in both eyes (no clarification is given regarding "deprived of sight"). In Egypt he is adjudged blind if he cannot see or count the fingers of one hand at a distance of one meter (l /60 or less). The only definition upon which all countries and investigators agree is abso-