In this paper, procedures and parameters that enable identification of, or discrimination between, general and localized corrosion processes through electrochemical noise are critically discussed. Their relevant similarities, differences, and interdependences are indicated. In addition, the relation between different procedures and parameters with the underlying physicochemical processes is indicated. Consistent and reliable information can be obtained from electrochemical noise data when a data analysis procedure is selected, which on the one hand has a high discrimination ability and on the other hand yields a descriptive parameter that is directly associated to the underlying physico-chemical process. Procedures that meet these two requirements appear to be the Hilbert-Huang transform (Hilbert spectrum), wavelet transform (energy distribution plot), and the analysis of charge and frequency of corrosion events. In addition to the procedures that meet these two requirements, some of the descriptive parameters can be determined by different analysis procedures, which increase their reliability.