The monosaccharide anhydrides levoglucosan, mannosan, and galactosan are known as ‘fire sugars’ as they are powerful proxies used to trace fire events. Despite their increasing use, their application is not completely understood, especially in the context of tracing past fire events using sediment samples. There are many uncertainties about fire sugar formation, partitioning, transport, complexation, and stability along all stages of the source-to-sink pathway. While these uncertainties exist, the efficacy of fire sugars as fire tracers remains limited. This study compared high-resolution fire sugar fluxes in freshwater sediment cores to known fire records in Tasmania, Australia. Past fire events correlated with fire sugar flux increases down-core, with the magnitude of the flux inversely proportional to the distance of the fires from the study site. For the first time, fire sugar ratios (levoglucosan/mannosan, L/M) in aerosols were compared with those in sediments from the same time-period. The L/M ratio in surface sediments (1.42–2.58) were significantly lower than in corresponding aerosols (5.08–15.62). We propose two hypotheses that may explain the lower average L/M of sediments. Firstly, the degradation rate of levoglucosan is higher than mannosan in the water column, sediment-water interface, and/or sediment. Secondly, the L/M ratio of non-atmospheric emissions during fires may be lower than that of atmospheric emissions from the same fire. Due to the uncertainties about transport partitioning (atmospheric versus non-atmospheric emissions) and fire sugar degradation along all stages of the source-to-sink pathway, we advise caution when inferring vegetation type (e.g. softwood, hardwood, or grasses) based purely on fire sugar ratios in sediments (e.g. L/M ratio). Future investigations are required to increase the efficacy of fire sugars as a complimentary, or standalone, fire tracer in sediments.