Several estimates of the cost of damage to building materials due to atmospheric attack have been produced. Methodologies employed usually fall into two broad categories, the simple mechanistic approach and the maintenance cycle approach. In the mechanistic approach physical damage is estimated from dose-response functions, and materials are assumed to be repaired or replaced when a critical damage level has been reached. The maintenance cycle approach also uses dose-response functions to estimate physical damage and considers how these will influence existing maintenance cycles for buildings. Neither approach deals adequately with the fact that atmospheric attack is not the only factor contributing to the need to repair or replace materials. Many other damage mechanisms can be involved in practice. This paper addresses the deficiency and presents arguments showing that inadequate treatment of the contribution of other damage mechanisms may lead to cost estimates for damage due to atmospheric attack which may be more than 60% too high. The paper suggests a plausible cost of damage to non-historic buildings within the UK arising from a typical UK coal fired power station of 0.07 p/kWh, within a range 0.04-0.19 p/kWh.