Lactose consumption has been associated with a high incidence of cataract in northern Indian and southern Italian populations. Galactose absorbed after hydrolysis of lactose from milk in individuals with normal lactase activity is considered responsible. However, lactase-deficient subjects who often avoid drinking milk are able to digest lactose and absorb free galactose in fermented milk and yogurt. This study was conducted to evaluate the relationships between milk and yogurt consumption, galactose metabolism and cataract risk. Milk ingestion was dose-related with cataract risk in lactose digesters (particularly in diabetics) but not in lactose maldigesters. Conversely, yogurt intake had a protective dose-effect on cataract formation for the whole population. Maximal galactose concentrations after an oral galactose test increased exponentially with age. Red blood cell galactokinase activity was significantly lower in elderly subjects (> 60 y) than in young individuals (P < 0.05), and galactose-1-phosphate uridyl-transferase activity was significantly lower in institutionalized subjects and in home-living elderly with cataract than in healthy elderly subjects (P < 0.05). We conclude that the cataractogenic action of milk lactose is dependent on the disturbance of galactose metabolism in elderly subjects and that yogurt is not cataractogenic, although the mechanism of the protective effect of yogurt remains unknown.