ABSTRACTThe rate of erosion of hillside gullies depends both on gully flow characteristics and the resistance offered by the gully soil profile to erosion. This paper describes a method for quantifying a physically‐based resistance measure, illustrated by application to a gully feeding sediment into the Bremer River, southeast Queensland, Australia. The dynamics of discharge down the gully during runoff events is the driver of erosion, but this was unknown. A new method is described whereby this unmeasured flow can be estimated using data on rainfall rate and river gauge monitoring.The data collected on the gully was the increase in dimensions and volume (and so soil loss) over a two year period. This information was obtained from a digital elevation model (DEM) of the catchment, derived from Light Detection and Ranging (LiDAR) observations made at either end of the two year period. The soil profile resistance characteristic evaluated is the energy required to erode a unit mass of soil from the gully walls, a physically‐defined parameter, J, present in flow‐driven erosion theory, which was adapted and applied to predict soil loss from the Bremer River gully. The value of J was evaluated by equating predicted to measured gully soil loss over the two year period using two alternative descriptions of gully cross‐section. Firstly a realistic gully shape description was used, made possible by LiDAR data, yielding J = 405.5 J/kg. Secondly, in order to allow use of more widely‐available aerial photography for such studies, the simplifying assumption of a semi‐circular gully shape was made, yielding J = 455 J/kg. Allowing a ±30% error in estimated effective runoff rate for this ungauged gully, the estimated J value would have an uncertainty of +1%/−7% using the actual gully geometry. The assumptions made in estimating J are discussed, and possible applications of this information listed. Copyright © 2014 John Wiley & Sons, Ltd.