This from the BBC news website yesterday (http://news.bbc.co.uk/) :
More schools to have drugs tests
Random drug testing of pupils is being introduced in secondary schools across Kent following the success of a pilot scheme in the county.
The Department for Education and Skills said it would evaluate the impact of the scheme, which is being implemented by Kent County Council (KCC).
A pilot scheme at Abbey School, in Faversham, saw pupils randomly selected and tested by taking mouth swabs.
The school said it was one factor which led to record GCSE results in 2005. Five good GCSE passes were achieved by 40% of pupils, compared with 26% in 2004 and 32% the year before.
During the scheme, 600 drug tests were carried out on pupils, chosen by a computer, at the specialist business and enterprise school.
The school's former headteacher Peter Walker said the fact only one pupil tested positive for cannabis showed the scheme had worked as a deterrent.
Now the government is to fund drug testing in several Kent schools.
After half-term secondary heads will be asked if they would like their school to sign up to the pilot scheme.
Parents will have to give their consent for children to be tested. At the Abbey School, consent was given for 86% of the 960 pupils.
"One of the benefits is that it helped children to resist peer pressure to join in with drug taking or other misuse," said KCC cabinet member for education John Simmonds.
"With all the external pressures on children we want to be able to support them as far as we can.
"If testing will help then it is worth pursuing."
Mr Walker was appointed as the UK Government's ambassador for random drug testing when he retired two months ago.
Earlier this month he was invited to Washington to discuss the scheme and met US drugs tsar John Walters at The White House.
Mr Walker said: "It is about time all schools came out and accepted that there is this problem that we need to address.
"Anybody who says they don't have a drug problem in their school isn't telling the truth." The government will commission research based on the forthcoming Kent pilot to establish whether there is a direct link between random testing and behaviour, attendance and academic achievement.
It's a bit of a leap of logic to suggest that this had any significant impact on improved exam results.