Fire severity is a key component of fire regimes, and understanding the factors affecting it is critical given the increasing incidence of wildfires globally. We quantified the factors affecting the severity of the 2019–2020 fires in Victoria, southeastern Australia. We constructed statistical models of relationships between fire severity (as reflected by two measures: Crown Burn and the composite measure of Crown Burn/Crown Scorch) and the main and interacting effects of five key covariates: fire progression zone (reflecting fire weather), time since previous major disturbance in the forest, forest type, slope, and aspect. The best supported models for the probability of a Crown Burn and the probability of a Crown Burn/Crown Scorch contained evidence of a three-way interaction between fire weather, forest type, and time since previous major disturbance as well as two-way interactions between (1) fire weather and slope, and (2) fire weather and aspect. There was an increase in the probability of Crown Burn and Crown Burn/Crown Scorch under more extreme fire weather in all forest types, with the effect especially elevated in dry forest. Our analyses also revealed a range of response curve shapes for the relationships between time since previous major disturbance and fire severity relationships and these varied by fire weather classes and forest type. Under severe fire weather conditions, we found that relationships between time since previous major disturbance and fire severity relationships often exhibited non-linear, negative polynomial shape with a peak around 10–40 yr, especially for Crown Burn, although there also were instances of this distinctive curve shape in our Crown Burn/Crown Scorch fire severity analysis. Our analyses also contained strong evidence that fire severity was higher on steeper slopes and on more exposed northerly aspects under extreme fire weather. Our analyses suggest that forests managed for timber production near settlements may be at increased risk of high-severity fire. This is because logging resets stand age to zero, after which there is a subsequent period of increased probability of high-severity fire, particularly under extreme fire weather conditions. Therefore, policies to maintain cover of older forest near settlements should be considered.