Laboratory exposures of copper have been performed at exposure conditions comparable to those in the UN ECE exposure program with respect to air flow conditions, relative humidity, and concentration of the gaseous pollutants sulfur dioxide and nitrogen dioxide. Extrapolation of the weight increases in the laboratory experiments match well those obtained at the test sites with high sulfur dioxide and nitrogen dioxide pollution levels. At these sites, sulfate and nitrate were the dominating surface constituents, as in the laboratory exposures. Additional constituents, detected in the laboratory, but not in the field, were sulfite and nitrite. At test sites with low pollution levels of sulfur dioxide and nitrogen dioxide, weight increases were much higher than in the laboratory exposures. At these sites, sulfate and nitrate were detected, but the relative amount of nitrate was much lower compared to the sites with high levels of sulfur dioxide and nitrogen dioxide. However, the amount of sulfate was practically the same. In addition, Cu2O was identified as an important compound at some sites. Characteristic of all sites with low pollution levels of nitrogen dioxide are the high levels of ozone. Ozone was not included in the laboratory experiments which could explain the discrepancy in weight increase. Laboratory experiments, investigating the combined effects of sulfur dioxide and ozone are presented in another publication. Both chloride and ammonia were detected as surface constituents on all field samples,confirming that salts and/or gaseous pollutants, other than sulfur dioxide, nitrogen dioxide and ozone, also are important for the understanding of atmospheric copper corrosion. However, the present investigation does not focus on these pollutants.