Samples of structural carbon steel were exposed to the atmosphere in a number of sites in Colombia and Cuba for periods of 2, 3, 6, 12 and 24 months. Corrosion rate was determined by mass loss measurements, and cause-effect relationship to exposure time and pollutants (deposited chloride and sulfur oxides) were calculated. Analysis let to adjust models by the multiple regression method. Also, comparison of atmospheric corrosion rates between Colombia and Cuba were done using covariance analysis.