Safety cameras have been central to strategy for improving road safety in the UK (and other countries) over the past decade and a National Safety Camera (NSC) programme was introduced in the year 2000 to fund and guide the expansion of camera sites. This programme was brought to an end in 2007 and responsibility for decisions on expanding, maintaining or rolling back safety cameras has now been largely devolved to local highway authorities and their partners. To make strategic choices in such areas it is clearly vital to understand the extent to which safety cameras are effective in reducing road accident potential. Although the NSC programme has been evaluated this work has focussed mainly on assessments of impacts at camera sites rather than wider area effects: particularly local highway area effects. The research reported here has explored the rates of personal injury collisions (PIC) over the past decade on the roads of English Local Authority areas and investigated relationships between area characteristics including levels of safety camera introductions and the PIC improvements (or otherwise) reported. To assist in this analysis, an indicator of safety camera intensity has been derived in this research to reflect the relative likelihood in areas of coming into contact with safety camera sites This work shows that area improvement rates differ significantly on an urban/rural dimension and are also very dependent on prior (to camera introductions) levels of PIC risk. Controlling for these factors using a pair comparison approach, analysis shows that safety camera introduction effects are statistically detectable in very highly urbanised, high PIC risk areas and at high risk sites within areas but not elsewhere. Whilst recognising that new high risk sites can emerge in any area, the conclusion is reached that many local areas and especially more rural areas would be unlikely to gain advantage from a further general expansion of fixed point and mobile camera sites and that de-commissioning of current sites should be seriously considered.