Many forest types are shaped by wildfire, which can reset ecological succession and modify species trajectories. Novel fire regimes are increasing globally, posing a major threat to forests. Understanding the successional dynamics of forests after wildfire is critical to develop evidence-based ‘benchmarks’ of forest structure to inform appropriate ecological management and restoration. However, this understanding is limited in intact forests over long periods, including those where old growth forest is sometimes used as a reference state. Several large, high-severity wildfires burned significant areas of the Mountain Ash (Eucalyptus regnans) forests in south-eastern Australia ≤1900, and in 1926/1939 and 2009. Here, we take advantage of this fire history to implement a natural experiment to better understand the patterns and environmental drivers of variation in forest structure at 84 sites across a multi-century chronosequence. We explored inter-specific differences across major plant lifeforms in relation to stem size and density, and their relative interactions over time. Our findings indicate that: 1. Early-successional forests ≤10 years after wildfire in 2009 are characterised by a high, but variable, density of homogeneous-sized stems (≥7000 stems ha^-1). In contrast, old-growth stands (≤1900 stand age) were characterised by the lowest number of stems, that were the least variable in density, but had the greatest variability in stem size, relative to other stand ages. 2. Environmental factors including slope and elevation influenced patterns of stem density, with the direction of the influence varying across lifeforms. 3. Interspecific relationships between individual plant lifeforms varied across the chronosequence, and were more negative in early-succession (2009 stand age), relative to late-succession.

Synthesis: Our findings demonstrate that rates of self-thinning in forests following wildfire (that occur within historical fire-regimes), may result in a ~50-60% decline in stem count in the several decades between early-successional forests and mature (1926/1939 stand age) forests, and then again between mature and old-growth forests. Self-thinning is a critical forest function, and represents one pathway through which structurally complex old-growth forests develop, providing significant ecological values. These empirical insights provide a benchmark to guide evidence-based restoration of montane forests in the face of predicted novel fire regimes.