This paper describes the peculiarities of carbon steel corrosion in very severe marine atmospheres and points out a number of uncertainties that still need to be explained, such as the following: (a) Data on the evolution of carbon steel corrosion (C) with exposure time (t) obeys the power function C ¼ At n . The variables upon which n depends are not fully known, though the salinity of the atmosphere undoubtedly plays a prominent role. (b) In marine atmospheres, the presence of akaganeite and magnetite phases among the corrosion products is especially significant. In relation with these corrosion products, a number of important questions remain unanswered: What conditions are necessary for their formation? Where are they preferentially located? Is there a critical atmospheric salinity concentration below they are unlikely to form? It is also of great interest to know the typical surface microscopic morphologies resulting from the presence of akaganeite, where the confusion among researchers is well known. (c) In very severe marine atmospheres, the morphology of the corrosion layers formed on steel can be highly unusual, such as sheet-type or mound rust. There is a lack of basic knowledge on the formation mechanisms and internal microstructure of these rust types.