The potential created when the metal comes into contact with a solution depends on both the metal and the solution. Two different metals or alloys in contact with the same environment will generally take two different potentials. If these two metals are electronically connected, the difference in their potential will give rise to electrochemical reactions and to the circulation of electrical current.

The most negative (least noble) metal is therefore positively polarized and the most positive metal is negatively polarized. In the vast majority of cases, this set-up corresponds to an increase in the corrosion rate of the most corrodible metal (the most negative), and a reduction in the corrosion rate of the least corrodible metal (the most positive). The severity of the corrosion of the least noble metal depends on several factors:

- The difference in potential between the two metals: The higher this value is, the more powerful the electromotive force of the phenomenon. The values to be taken into consideration correspond to the potentials of metals and alloys constituting a pair in relation to the environment in question. These potentials are experimental in size, and must be distinguished from the standard potentials of thermodynamics tables. The experimental potentials are strongly influenced by such parameters as temperature, agitation and aeration. Moreover, some metals can take two different potentials in relation to the same environment, depending on whether they are active or passive (stainless steel coming into contact with seawater, for example). These considerations demonstrate that it can be difficult to foresee trends without having recourse to experimentation, because many parameters are liable to reverse the polarities of certain galvanic pairs. The table opposite shows the galvanic pairs of the principal metals.

Example of galvanic corrosion for a riveted assembly of aluminium and copper sheets, without insulation. There is corrosion, because the dissolution potential of the aluminium is lower than that of copper.

- The surface ratio between the two metals: the least favourable case is that of a large cathodic surface (the most positive metal) electrically connected to a small anodic surface (the most negative metal). The corrosion rate of the most negative metal can be multiplied by 100, or even 1000. An assembly consisting of sheets of iron held by copper rivets is far more resistant to corrosion than the opposite set-up (copper sheets and iron rivets).

- The conductivity of the corrosive environment influences the location of deterioration. Galvanic corrosion can occur even in very resistant environments. In such cases, it happens at the point where the two metals touch. Conversely, the attack is located less in the conductive environment.

- The corrosion resistance of the most noble metal- regardless of its potential - has considerable influence on the behaviour of bimetallic torque. If the noblest metal corrodes, its corrosion products risk, through their movement, accelerating the corrosion of the most corrodible metal. For example, the products of copper corrosion can corrode aluminium. The result is that the copper-aluminium pair (which has a lesser difference in potential than the gold-aluminium pair) is, however more dangerous – because gold, which is incorrodible, does not present this risk.

A few ways of fighting against this corrosion: choose metallic pairs in which the elements are as close as possible in the table opposite; avoid an unfavourable surface relationship; avoid, as far as possible, direct contact between two different materials (by using a seal, insulator, coating, etc.).