AbstractMany corrosion phenomena are nonuniform, which means that anodic and cathodic locations are spatially separated. An example is macrocell corrosion of steel in concrete. Under these conditions, determining the corrosion rate from polarization resistance measurements and using the Stern–Geary equation is fundamentally not possible. We present a novel theoretical approach for the interpretation of galvanostatic pulse measurements, to make them applicable as a method for corrosion rate measurements in situations of localized corrosion. Experiments show that it is important to consider that (a) only a fraction of the applied current flows through the anode of the macrocell, and (b) this current is not constant over time. We propose an approach to quantify and consider these two effects, based on information generally accessible in condition assessment of concrete structures. Our results show that galvanostatic pulse measurements are a robust method to determine the corrosion current. With the traditional empirical approach, the measurement error was generally below factor 3, and occasionally up to factor 10. With the novel approach, this error could be reduced to a factor of maximum 2 in all cases.