The main cause of corrosion initiation in steel reinforcement within concrete is the presence of chlorides above a threshold value. The primary source of chlorides is from the external environment, either from marine structures or de-icing salts. The most commonly used method of predicting or modelling chloride ingress is to use Fick's second law. Inevitably, there are significant variations involved in the positioning of the steel, the exposure conditions surrounding the concrete and the concrete itself. Due to the variability in chloride penetration and cover it is commonly found that the data will suggest that somewhere on the structure, chlorides have reached the depth of the steel. A single reinforcement bar with very low cover is, however, simple to address. It is more relevant to know when a significant proportion of the reinforcement is likely to become exposed to higher chloride concentrations. As structures are required to last for longer periods, and existing structures are reused for new purposes, the prediction of the likelihood of corrosion becomes more and more important. This paper presents two methodologies that may be applied to obtain a statistical assessment of chloride penetration from site data.