Clonal Eucalyptus plantations in the Pointe-Noire region have been established on sandy and nutrient poor soils, and carry potential risk of depleting water and nutrient. To assess these risks, water and nutrient budgets should be calculated. However, the accuracy of hydrological models is strongly dependent on a realistic description of root distribution in soil. The spatial distribution of root systems in a chronosequence of clonal stands was then studied, using root intersect counting. The stands were 3 months, 6 months, 1 year, 2 years and 9 years old. For each stand we studied three vertical profiles perpendicular to the planting row at different spacing from a representative tree. The profiles were divided into square grid cells of 25 cm2, and the number of roots belonging to three diameter classes (between 0.1 and 1, 1 and 10 mm, and over 10 mm) were counted in each grid cell. The profiles were 2.50 m wide (half the distance between the planting rows) and 3 m deep, except for the two youngest stands (1.5 m deep). Spatial statistical analyses and analysis of variance were carried out to describe root distributions and their spatial and temporal changes. We observed that: The trees quickly developed an extensive root system. One year after planting the root system extended to depth beyond 3 m deep and up to the middle of the inter-row. Root density decreased sharply with depth, with most fine roots in the surface layers. The number of fine root intersects in the 0-25 cm surface soil layer represented 16-53% of the total throughout the profile depending on stand age and the type of profile. The percentage of root intersects in surface layers increased with stand age: this may reflect the greater concentration of nutrients in surface layers. High variability in root densities was observed in the surface soil layers but no gradient of decrease from the planting row to the middle of the inter-row could be underscored. However, the highest root densities were genera