Spatial patterns of seedling distribution and diversity were analysed in small fragments of holly Ilex aquifolium L. woodlands and in their surrounding areas. Two sampling locations with similar structure were selected for this study: Oncala and Robregordo. They consist of nearly monospecific Ilex stands surrounded by grasslands with high scrub abundance. The seedling appearance of woody species was quantified from March to November 1998. Sampled areas were: 1) closed holly canopy; 2) open holly canopy or small forest gaps; 3) holly woodland edge; 4) surrounding grassland; 5) under isolated fleshy-fruited shrubs scattered over the grassland; 6) under dry-fruited shrubs and 7) the closest forest to the holly woodland. Additionally, a pine forest at a distance of 20 km from Oncala was sampled. In every area ten permanent 50 × 50 cm quadrats were fixed for monthly seedling control. The highest germination density occurs under the holly woodland, especially in closed canopy areas. Nevertheless, these closed woodlands neither maintain a great quantity of surviving seedlings nor a high diversity. Seedling density is considerable in canopy gaps, shrubs and forest edge, and these habitats have greater diversity values than understorey habitats. Fleshy-fruited shrubs maintain higher seedling densities and diversity than dry-fruited shrubs. Woody seedlings are rare over the grassland. The three non-holly forests studied have very similar seedling densities and diversity values, higher than those under closed-canopy holly. Regional differences are important for the numbers of seedlings surviving from previous years, which are scareer in Robregordo. However, little difference is observed in spatial patterns of seedling diversity between the two locations. We discuss a number of processes affecting seed rain density and differential mortality rates that could account for these spatial patterns, namely competition for light inside the woodlands, facilitation under nurse shrubs, or competition escape at the edge. Environmental factors such as moisture stress may play a role in regional differences.