Geodemographic systems currently provide classifications of small areas based primarily on their demographic and socioeconomic characteristics. This paper presents an innovative use of geodemographic methods in the public sector, in particular to determine people's road injury risk. It uses Stats19 road accident data from the London Metropolitan Police, specifically postcode information relating to the driver and casualty of each road accident from 1999 to 2003. Experian's Mosaic software was used to append each postcode to a geodemographic Mosaic type. The results showed distinct spatial and statistical prevalence of certain groups within society to be more at risk of being in a road accident. This highlights the need for a more sophisticated understanding of the different risks people face as they engage in the road environment.