Dead fuel moisture influences the risk of fire ignition events, with implications for fire hazards, risk mitigation, and the design of prescribed burning activities. Because direct fuel moisture measurements are rarely available, fuel moisture must be estimated when evaluating fire risks. Most estimates rely primarily on atmospheric conditions and ignore the interaction of fuels with the soil surface with which they are in hydraulic contact. In this study we explore whether dead fuel moisture predictions can be improved with information about surface soil moisture. Despite the likelihood that dead fuels would exchange water with underlying soil, the influence of soil moisture on fuel moisture has been poorly studied. An analysis of 202 observations of co-located soil moisture, 1 and 10 h fuel moisture measurements, with environmental and meteorological covariates, showed that soil moisture had a small but significant effect on fuel moisture across all sampled conditions. The influence of soil moisture on fuel moisture was the most important among all other environmental factors for wet soil conditions and 10-h fuels, where a 1% increase in soil moisture content led to approximately a 0.6% increase in fuel moisture. The effect of soil moisture on 1-h fuel moisture although significant was small. Incorporating spatio-temporal variability of soil moisture into time-series or spatial predictions of fuel moisture tended to (i) reduce the seasonal duration in which fuels had a high probability of ignition by an average of 52–60 days across the tested years, and (ii) increased the heterogeneity of the probability of ignition through space, compared with similar models that did not incorporate soil moisture information.