CRUCIAL new evidence from two of Britain's most distinguished medical experts could provide the key to freedom for one of Scotland's longest serving prisoners.

Andrew Smith was jailed for murder 21 years ago after a jury was told that his victim probably died from a kick to the head.

The fresh evidence strongly suggests that the head injury was ''far more likely'' to have been caused by a fall than a kick and that there is nothing to indicate that the fractured skull suffered by the victim was inflicted by a boot to the head.

Smith's Glasgow solicitor John Macaulay is confident that the new evidence is powerful enough to persuade Scottish Secretary Donald Dewar to refer the case back to the Court of Criminal Appeal.

He told The Herald: ''I think this is an absolutely appalling case. The conviction in 1977 was bad enough but what has happened since is infinitely worse.''

Smith, who is being held at Greenock prison, was convicted of murdering Richard Cunningham, 29, following an incident in a bar in Larkhall.

The prosecution case was that the victim died as a result of a kick to the side of the head, a claim which was based on evidence given at the trial by the late Professor Arthur Harland, the leading forensic scientist of his day. Professor Harland's evidence that a kick was the most likely cause of the victim's skull fracture was supported by Dr Alan Watson.

Smith, who was 18 at the time and a keen amateur boxer, insisted that he had been acting in self defence after being attacked, had punched Mr Cunningham just once, and that he died as a result of striking his head in a fall, possibly some time later.

Lord Avonside, the trial judge, did not allow the jury to consider the possibility of the lesser charge of culpable homicide and Smith was convicted at the High Court in Glasgow by a majority verdict - he claims it was 8-7 - leading to a mandatory sentence of life imprisonment.

Smith was released after six years, recalled after a relatively minor offence in 1985, released again that year, but after 13 months his freedom was again terminated in November 1986. He has spent about 19 of the last 22 years in custody.

Smith's main problem when he was allowed out seemed in handling drink, but Mr Macaulay insists: ''He is not a violent man and his licence should never have been recalled for one or two very trivial things which cannot remotely justify what has happened since.''

The new evidence has been provided by Professor Bernard Knight CBE, a barrister and a consultant in forensic medicine and pathology, and Dr Helen Whitwell, a consultant forensic pathologist and neuropathologist based at the Queen Elizabeth Medical Centre in Birmingham.

Professor Knight points out that Professor Harland was neutral about the cause of the head injury, saying only that the skull fracture resulted from ''an impact'' on the left side of the head.

''Where I strongly disagree with Professor Harland is in his interpretation of the mechanism of the head injury, which he attributes in his trial evidence to a kick to the head. He totally ignores the unequivocal presence of 'contre-coup' brain damage.

''My opinion is that the head injury was far more likely to have been caused by a fall than a kick.''

Contre-coup injury, remote from the site of impact, is a reliable indicator of deceleration injury, the moving head being drastically arrested on contact with a fixed surface. It is the classical way of distinguishing a blow from a fall.

In her report Dr Whitwell states: ''In my opinion there is no evidence to support kicking as a mechanism of the skull fracture and the findings of this fracture, together with the head injury, are consistent with occurring in a fall.''