Multiple regression analysis, with lags, was performed on daily mortality on air pollution and weather variables for three United States cities (New York, Philadelphia, Los Angeles). Relationships between mortality and air pollution weather were found for all three cities. These relationships were consistent within cities, by season. They were also consistent between cities, allowing for generalization about the relationship: daily mortality is associated with air pollution and weather in urban areas of the U. S. This association can be stated as a formal relationship in time. The effect of pollution and weather on mortality is essentially an immediate one. (Author)