In a chronosequence of afforestations adjacent to ancient forest, relations between herbaceous forest vegetation and possible explanatory variables, such as soil characteristics, land-use history variables and relative insolation on the forest floor, were studied using linear models on species groups and a direct gradient analysis (CCA). An explorative correlation analysis indicated that the moisture, carbon, total nitrogen and phosphorus contents of the soil, were correlated with forest age. Although soil pH was not correlated with forest age, its variability was greater in ancient forest and 19th century afforestations than in recently afforested parcels. The linear models indicated that soil pH positively influenced the diversity and cover of both slow-colonizing and fast-colonizing woodland species. P enrichment, in combination with a high relative insolation, stimulated Urtica dioica in recent afforestations. Deep shadow on the forest floor suppressed light-demanding species, such as U. dioica , and indirectly favored woodland species. The multivariate analysis confirmed that floristic diversity was best accounted for by the variables plant-available P, pH (KCl), nitrate (negatively correlated with insolation) and forest age. The discussion focuses on the origin of the observed variability in soil characteristics, natural or anthropogenic, and the impact on vegetation succession. It is concluded that most woodland species can colonize recently afforested farmland parcels in Muizen forest, as soil pH is not a limiting factor. However, P eutrophication can have an indirect negative effect, increasing competitive exclusion by U. dioica . Plantation of an understorey can suppress U. dioica and stimulate colonization of shade-tolerant woodland species in afforested farmland.