Studies about Pb variability in contaminated soils, especially in the case of fallout from smelters or facilities, are few and the relevance of geostatistical technique for studying contaminated soil as well as the surface area required for investigation are unclear. Few studies have been conducted on short-range variability of trace elements in soils, and some of them have shown these elements to be randomly distributed while others pointed out the spatial dependence between sample locations. Therefore the main objective of this study is to document the short-range variability of Pb in an area contaminated by fallout, using a case study in France. Soil lead in this site arises from a history of industrial use. No other metals were detected in Pb emissions. On this site, we conducted a statistical and a geostatistical study of Pb content in soil, and of soil characteristics that may affect it. The aims of this work are (i) to separate the spatial structure due to high Pb fallout from a content and diffuse pollution, (ii) to separate the variability due to analytical measurements from ones related to spatial variations, (iii) to point out the soil characteristics correlated with Pb, (iv) to optimize sampling design for monitoring the long-term changes following remedial action. Results demonstrated that in this highly contaminated soil, fallout Pb was mainly located in the surface layer (0–0.2 m depth). The geostatistical study suggested that the study of the spatial structure may help in discriminating fallout Pb from Pb coming from diffuse pollution or geochemical content. However, the large spatial variability of Pb, partly due to microheterogeneity made the monitoring of this toxic element very difficult, and required a larger surface for sampling. Using correlation techniques to detect causations appeared not to be relevant. Therefore, other statistical techniques such as path analysis should be tested. Additionally, microanalysis, mobility and bioavailability studies, as well as lab or bench scale remediation studies should be conducted. Key words: Contaminated soils, fallout Pb, short-range variability, geostatistics