Tropical savannas are characterized by high primary productivity and high fire frequency, such that much of the carbon captured by vegetation is rapidly returned to the atmosphere. Hence, there have been suggestions that management‐driven reductions in savanna fire frequency and/or severity could significantly reduce greenhouse gas emissions and sequester carbon in tree biomass. However, a key knowledge gap is the extent to which savanna tree biomass will respond to modest shifts in fire regimes due to plausible, large‐scale management interventions. Here, we: (1) characterize relationships between the frequency and severity of fires and key demographic rates of savanna trees, based on long‐term observations in vegetation monitoring plots across northern Australia; (2) use these relationships to develop a process‐explicit demographic model describing the effects of fire on savanna tree populations; and (3) use the demographic model to address the question: to what extent is it feasible, through the strategic application of prescribed burning, to increase tree biomass in Australian tropical savannas? Our long‐term tree monitoring dataset included observations of 12,344 tagged trees in 236 plots, monitored for between 3 and 24 years. Analysis of this dataset showed that frequent high‐severity fires significantly reduced savanna tree recruitment, survival, and growth. Our demographic model suggested that: (1) despite the negative effects of frequent high‐severity fires on demographic rates, savanna tree biomass appears to be suppressed by only a relatively small amount by contemporary fire regimes, characterized by a mix of low‐ to high‐severity fires; and (2) plausible, management‐driven reductions in the frequency of high‐severity fires are likely to lead to increases in tree biomass of about 11.0 t DM ha−1 (95% CI: −1.2–20.8) over a century. Accounting for this increase in carbon storage could generate significant carbon credits, worth, on average, three times those generated annually by current greenhouse gas (methane and nitrous oxide) abatement projects, and has the potential to significantly increase the economic viability of fire/carbon projects, thereby promoting ecologically sustainable management of tropical savannas in Australia and elsewhere. This growing industry has the potential to bring much‐needed economic activity to savanna landscapes, without compromising important natural and cultural values.