Every police hunt for a dangerous criminal now seems to have a psychologist on hand to give a rundown of the most likely suspect. Even television drama abandoned the supersleuth recently for Robbie's Coltrane's memorable gambling, shambling, heavy- drinkingpsychologist in the series Cracker.

So has good old-fashioned detective work become a quaint irrelevance? Absolutely not, according to David Canter, the man widely credited with pioneering modern criminal psychological profiling in Britain. 'Offender profiles do not solve crimes, only good police work does that,' he says. 'It just happens that it is an area the media are mesmerised by at the moment and they've turned it into something exotic.'

So fed up is Professor Canter with what he calls all this 'hit- and-run psycho-babble', he intends to set the record straight in his new book Criminal Shadows. The purpose of profiling, he explains, is to do little more than help police to prioritise the many suspects that they have to sift through in the course of a criminal inquiry. It can also point investigators in the direction of certain kinds of individuals as possible suspects who might not otherwise have been considered. No more, no less.

Ironically, it was his own success in a police investigation that gave so much weight to offender profiling. His work on the 'railway rapist' case in 1986 broke new ground and led to the conviction of John Duffy for a series of rapes in the London area, culminating in two murders.

The professor had been approached in 1985 by two senior Metropolitan Police officers to discuss the feasibility of behavioural psychology helping them in their work. When he finally embarked on the Duffy case, he knew he was following in the footsteps of the FBI's Behavioural Science Unit - made famous by Silence of the Lambs.

'But neither Silence of the Lambs,' recalls Canter, 'nor the publications and lectures of the FBI behavioural science agents indicated in any detail just how to go about producing an 'offender profile'.'

Using general psychological theories, Professor Canter, with the help of Rupert Heritage, of Surrey Police, began organising data of the man's attacks under headings such as days and times of attacks, locations and so on. Other behavioural details were added, until certain speculations about the man's methods and character could be made. 'The accuracy of the profile surprised everyone, especially me,' he says.

On the 'railway rapist' suspect list, Duffy already ranked 1,505 out of around 2,000 men, due to his blood type, but because he matched the profile so well, the police decided to mount a large-scale surveillance to watch him. From this operation enough evidence was gathered to make an arrest.

'I never thought,' says Professor Canter, 'that this first attempt at a profile would become of such interest. In academic terms, it was merely the equivalent of a pilot project. Since then, we've actually been trying to pin down the processes that gave rise to its success.'

It is not just the public that needs to temper its enthusiasm. The police themselves need to understand that offender profiling is more often an art than a science, believes the professor.

This is the prime reason for the establishment last year of his postgraduate course at Surrey University called 'investigative psychology' (a term he coined), which has in effect established an entirely new branch of psychology. It is the first time that police and psychologists are studying together the ins and outs of criminal behaviour.

While Professor Canter's approach has some strong support among policemen, it has to be said that there are others who are less than happy. 'There are a number of entrenched police officers who want a tame expert who will give them the opinions that they are seeking,' he says. 'I believe it is absolutely crucial to have investigative procedures and training that take in a much broader range of issues than simply, 'What are the characteristics of the offender we are looking for?' It needs to be recognised that profilers can get it seriously wrong.'

What better warning than the disastrous Waco siege, with the array of America's finest FBI officers and other psychiatric profilers on hand to guide and advise, says Professor Canter. 'The consensus was that David Koresh and his followers would give up once tear gas was pumped into their compound. Yet it appears that they set fire to the place and then shot themselves. So, in fact, these expert profilers got it completely wrong.'

'Criminal Shadows' is published today by HarperCollins, pounds 16.99.

(Photograph omitted)