Speed camera traps don't always cut accidents, says a survey in Britain.
Data shows that collisions increased at a number of locations after the introduction of cameras.
In one area, the number of accidents actually went up at 17 of its 89 speed camera sites.
Seven of the 44 speed trap sites in the Thames Valley - an area which follows the River Thames from Oxfordshire in the west to London in the east - saw a hike in the number of collisions, while four of the 47 locations north of the city also saw an increase in accidents.
The statistics were revealed after Britain's Department for Transport called on local authorities to publish accident and casualty rates for camera sites before and after the introduction of speed cameras.
Road Safety Minister Mike Penning has billed the move as an opportunity for residents to hold councils to account and ensure ineffective cameras are switched off.
But although the department of transport set a July 20 deadline for all English councils to publish their speed camera statistics, almost half have failed to do so, leaving their residents in the dark as to how effective road safety measures in their areas really are.
The department of transport has said the remaining 72 authorities will make their figures public in the next few weeks.
RAC motoring strategist Adrian Tink said: "This information is long overdue. We've always said there needed to be more transparency on speed camera performance.
"But local authorities need to actually use this data and investigate speed camera sites that aren't working.
"If cameras are not cutting accidents, they need to be replaced with measures that do work."
According to Tink, there are around 6000 speed camera sites in Britain, generating around $180 million a year in fines.