For economic reasons, many large spillways discharge into unlined rock channels or into the river bed directly. The acceptability of such designs is based on engineering assessment that the unlined channel is sited in rock judged to be sufficiently resistant to erosion at least for high frequency floods. It is accepted that under larger low frequency floods there may be some erosion and damage to these unlined channels, but the damage must be limited and must not endanger the structural integrity of the dam.However, in recent years in Australia there have been some unlined spillways which have experienced significantly greater erosion than was initially anticipated and have required substantial remedial works costing tens of millions of dollars.Many of the methods currently used to assess rock erosion on steep slopes are developed based on limited experimental data and involve significant level of uncertainly [Annandale, 1995, 2006]. There has also been recent development of process models that incorporate fluid-rock matrix mechanical coupling effects [e.g. for plunge pools, Bollaert, 2004]. It is well established that aeration significantly changes the drag characteristics of supercritical flow on steep concretelined spillways [Wood, 1991; Chanson, 2004].The purpose of this present contribution is to evaluate the potential significance of aeration in determining drag levels, and therefore influencing rock detachment processes on unlined spillways. This has been undertaken using conceptual characterisations of aeration have been developed across hydraulic engineering.Specifically, we summarise present understanding of the inception, behaviour and implications of aerated flow down steep, rock-lined slopes and the consequences for scour protection design. We also critically review numerical techniques presently being developed and applied to this problem.The most significant aspect of this contribution is a successful reconciliation of recent conceptual microphysical characterisations of aerated flow with conventional engineering approaches based on dimensional analysis of large-scale test data. Further work will be directed at reconciling recently-developed mechanical coupling approaches for unlined spillways with analyses for aerated flow over placed rock in steep channels.