In forests subject to stand-replacing disturbances, early successional stands can provide important habitats for a range of species not typically present in long-undisturbed areas. Compared to old-growth forests, the habitat values of – and key ecological processes in – early successional forests have been less studied, perhaps due to a perception that early successional forests revert to a homogenous “clean slate” following stand-replacing disturbances. In this paper, we draw on 36 years of long-term research in the Mountain Ash (Eucalyptus regnans) and Alpine Ash (Eucalyptus delegatensis) forests of south-eastern Australia, together with examples from elsewhere around the world, to show that not all kinds of early successional forests are created equal. We argue that the ecological values of early successional forests can be profoundly affected by six inter-related factors: (1) The evolutionary context and environmental domain of a given ecosystem. (2) Successional stage and condition of a forest stand prior to disturbance. (3) Disturbance intensity, severity and type (e.g. wildfire versus clearcutting). (4) Post-disturbance conditions including climate and weather. (5) Post-disturbance management (e.g. salvage logging) which can have significant impacts on biological legacies. And, (6) The relative spatial extent and spatial arrangement of early and late successional forest across a landscape. These factors can influence ecological values directly, or through effects on the types, amount and spatial patterns of biological legacies present in early successional forest. We present a conceptual model highlighting the inter-relationships between these factors and illustrate its use through a detailed case study. Strategies to improve the management of early successional forests include: (1) Identifying the species associated with post-disturbance environments and the reasons why they occur in such environments. (2) Understanding the types, numbers, and spatial patterns of biological legacies that remain after natural disturbance. (3) Identifying critical areas that should be excluded from logging or other human disturbance. (4) Limiting the extent of post-disturbance activities like salvage logging that undermine the ecological values of, and ecosystem processes in, early successional forests. And, (5) Balancing the relative amounts of early successional versus late successional forest in a given landscape or region to ensure that a variety of forest types are present at any given time, and that critical biological legacies are retained. Paradoxically, ensuring that landscapes support extensive areas of late successional forest is critical so that future early successional forests are not devoid of the biological legacies necessary for ecosystem function and recovery.