Both large-scale physical infrastructure redesign and extensive use of in- vehicle driving assistance systems can contribute to improving road traffic safety. Limited availability of effect data (historical and estimated) for both alternatives is hampering long-term strategic analysis of their potential effects. This paper investigates the use of a first-order and one- variable grey model, denoted as GM (1,1), to forecast the trend of the reduction of traffic accident severity (in terms of fatalities and hospitalisations) through mentioned strategies and combinations thereof. Based on modelling the limited available data of the effects of the infrastructure redesign programme in The Netherlands for the period 1998-2002, we forecast the trend of fatalities and hospitalisations for the years 2003 until 2010. The result is compared with other traffic safety enhancement scenarios by using cost-effectiveness analysis (CEA). Error analysis shows that the applied model has a high degree of reliability. Therefore, the method (grey model and CEA) and the outcome of the analysis may contribute to planning and decision making concerning further appropriate steps to reach the ambitious Dutch road traffic safety goals for 2010.