The influence of topsoil and fertilizer application on denuded road construction sites was evaluated to assess its contribution to improvement of vegetation re-establishment. The study sites were within a mixed hardwood and conifer forest on crushed, unweathered subsurface material with low fertility and low biological activity. Topsoils were removed from the site, stockpiled and reapplied to the site after construction. The effect of topsoil amendment on plant growth, soil fertility, mycorrhizal infection and an index of microbial biomass were measured in field and greenhouse experiments. Plant growth on the topsoil amended field plots were greatly increased relative to treatments with fertilizer but no topsoil. Three years after establishment, dry weight production on the plots without topsoil treatment was about 40 per cent of the plots treated with topsoil. Greenhouse experiments were designed to compare fresh, dried and stockpiled topsoil. These experiments indicated that storage of the harvested topsoil for five months in a stockpile had minor effects on plant growth, soil fertility, mycorrhizal infection and microbial biomass. Topsoil volumes had to exceed 20 per cent of the total soil volume to achieve statistically significant benefits and higher ratios showed greater benefit. The percentage of mycorrhizal infection was greatest in topsoil treatments without fertilizer. The addition of fertilizer increased growth but reduced the percentage of roots forming mycorrhizae. When the total weight of infected roots was calculated, however, infection was found to be greatest with a moderate level of fertilizer (equivalent to approximately 27 kg N ha-' and 39 kg P ha-'), and was less in both higher fertilizer treatments and in unfertilized treatments. Topsoil amendment increased microbial biomass nitrogen but fertilizer treatment did not.