The distribution of the galvanic corrosion of a nickel‐to‐steel couple in open, slowly moving, tap water was estimated from measurements of potential differences in the water around the couple. The distribution was confirmed later by thickness measurements. The two methods were in excellent agreement. The maximum galvanic corrosion of the steel occurred at the junction of the two metals and decreased as the distance from the junction increased. The maximum effect at the junction was 3 1/2 times the average galvanic corrosion. The normal uncoupled corrosion of the steel was greater than the average increased corrosion due to the galvanic effect despite the three‐to‐one area ratio of nickel to steel. Potential measurements can be made as soon as a steady state is reached, but it usually takes considerable time for corrosion to become extensive enough for thickness measurements to be accurate.