In assessing the treatment of sites with low skid resistance, the relative priority of lengths marginally below skid resistance threshold where there have been a number of recent crashes must be balanced with that of lengths substantially below the threshold with no recent crash history. The judgment is complicated by the highway authority having a duty of care to maintain the road in a safe and serviceable condition, so it is not acceptable to let the skid resistance deteriorate indefinitely, even if there have been no crashes. Furthermore, the analysis lengths are generally short and the number of crashes is generally small, leading to a high degree of uncertainty in estimating the underlying risk. And finally, as there can be a large number of sites that require investigation within any particular jurisdiction, there is a need for a simple method otherwise the whole process of site investigation demands a disproportionate level of staff resource. This paper describes an accident risk model that has been created to provide a simple and consistent method for rating the priority of treatments at locations with low skid resistance and the dilemma for determining how to incorporate crash data within decisions on treatment priority.