Galvanized steels have been extensively used in auto industry, construction and household appliances etc. due to their excellent performance in corrosive atmospheric environments. Substantially, the degradation of galvanized steels could be divided into four stages: τ1 : dissolution of zinc, τ2 : sacrificial corrosion of zinc by coupling with steel, τ3 : corrosion protection by zinc corrosion products, and eventually τ4 : general corrosion of underlying steel. Lots of works have investigated the corrosion behavior of this type of material. However, there are very few works on the corrosion protection mechanism after zinc coating exhausted. Therefore, the present work focused on the corrosion stage τ3 during which the underlying steel is merely covered and protected by zinc corrosion products. It is essential to evaluate such a stage to predict the service life of galvanized steels more precisely. Based on lab experiments, the applicability of the electrochemical measurements as non-destructive evaluation methods was investigated as well. In this study, commercial electro-galvanized (EG) steel and hot-dip galvanized steel (GI) were used. For EG, the thickness of zinc coating was 3 μm, 5 μm and 10 μm, and was about 10 μm in average for GI. The samples, which were deposited with 5 g/m2 and 2 g/m2 NaCl namely, were exposed to the dry-wet cycling of 95% RH-2 hour and 40% RH-2 hour in a chamber. The corrosion process was monitored by measuring the corrosion potential and electrochemical impedance spectroscopy at some interval, and the remained amount of zinc was quantified by the anodic stripping method. In addition, the cross-section was observed by SEM, and the composition was analyzed to identify the corrosion products. Taking the results of 3 μm EG as an example, basically the corrosion potential did not rise directly right after the zinc coating disappeared. Characteristic impedance value also indicated a stage with good corrosion resistance after the zinc coating exhausted. The high frequency impedance values (ZH), which represented the sum of corrosion products and solution resistance, nearly kept stable during this period of time. Meanwhile, the reciprocal of low frequency impedance values (1/ZL) represented the low corrosion rate, which confirmed the protectiveness of zinc corrosion products. The zinc corrosion products strongly suppressed mass transports of oxygen and dissolved ferric ions. And such mechanism was proved by the cross-sectional view of the samples at τ3 stage, showing that a compact layer of simonkolleite attached to the steel surface. Finally, 1/ZL increased gradually due to the onset of the substrate steel corrosion (τ4 ), when the adhesive zinc corrosion products disappeared. As further discussion, some rust points were observed on the sample in τ3 stage during the 5 g/m2 high concentration lab test. It was supposed that the distribution of zinc corrosion product was uneven, particularly for thin zinc coating. In contrast, less rust point was observed on the samples with thicker zinc coating during the τ3 stage, and the protective effect of zinc corrosion products lasted much longer. Besides, the GI showed better performance than EG, which might be due to the crystal structure of zinc and the presence of alloy layer. Additionally, with less amount of deposited salt during the chamber test, such lab test procedure could simulate the sluggish corrosion behavior in real atmospheric environment. To conclude, by applying the non-destructive electrochemical impedance spectroscopy, the degradation stages of zinc are well distinguished. Especially the presence of τ3 stage was identified, where the underling steel was well protected by zinc corrosion products.