The ability of inorganic geochemistry to record environmental change and especially human impact has been evidenced by several studies across Europe, especially in peat, where it is possible to record the impact of agriculture, mining and other industries. However, despite the numerous investigations on the impact of ancient human activities such as ore mining and smelting, little attention has been paid to geochemistry as a tool to solve problems of palaeopollution in the surroundings of archaeological sites. This paper presents geochemical evidence of the impact of a possible early Roman road built in SE Belgian peatland. Increased Zn and Pb concentrations suggest that Pb–Zn ores were transported on the road. Lead isotope analyses suggest that these ores are locally derived, being compatible with those found in the nearby Pb–Zn ore deposits from East Belgium. Present results provide direct evidence that East Belgian Pb–Zn ores were already being mined during Roman times, i.e. earlier than previously suspected (i.e. 14th century) and that Zn appears to be relatively immobile here. On a broader scale, it also demonstrates that such an early road already had an impact on the environment in terms of metal pollution. This paper enlarges on the range of possibilities offered by geochemistry in the field of geoarchaeology.