We review the approaches for estimating biomass burning from ice-cores and consider the challenges and assumptions in their application. In particular, we consider the potential of biomarker proxies for biomass burning, hitherto not widely applied to glacial ice archives. We also review the available records of biomass burning in ice-cores and consider how variations in fire regimes have been related to atmospheric and land-use changes. Finally, we suggest that future developments in ice-core science should aim to combine multiple biomarkers with other records (black carbon, charcoal) and models to discern the types of material being burnt (C₃ versus C₄ plants, angiosperms, gymnosperms, peat fires, etc.) and to improve constraints on source areas of biomass burning. An ultimate goal is to compare the biomass burning record from ice-cores with hindcasts from models to project how future climate change will influence biomass burning and, inversely, how fire will affect climate.