Because of excessive reliance on the method of ecological correlation, in which the units of study are entire communities, there have been few agreed conclusions concerning the relation between mineral quality of water and health. However, in recent years there have been attempts to relate water exposure to health outcome in statistical series of individual subjects. Studies in Kitchener, Ontario and Regina, Saskatchewan found that households having domestic water softeners experience lower death rates than others. Kitchener data clearly implied an association between copper piping and mortality rate but this was not confirmed in Regina. Comparison of myocardial tissue between residents of soft and hard water areas has confirmed that insufficient magnesium intake is a likely cause of higher mortality in soft water areas. Within the 'normal' range of myocardial magnesium concentration the risk of IHD death appears to vary by at least one and possibility two orders of magnitude.