Professor Sir Richard Doll was granted yesterday what few medical scientists live to see - the fruits of their research measured in hard figures showing millions of lives saved.

Professor Sir Richard Doll was granted yesterday what few medical scientists live to see - the fruits of their research measured in hard figures showing millions of lives saved.

Fifty years ago next month, the Oxford University cancer epidemiologist, one of the world's most distinguished tobacco researchers, published his first paper in the British Medical Journal establishing the link between smoking and lung cancer.

Yesterday, aged 87, he presented the results of a follow-up study that showed deaths from lung cancer were falling faster in Britain than anywhere else in the world.

Because of that early warning, published in the BMJ on 30 September 1950, and the studies and campaigns that followed it, the terrible carnage wreaked by smoking has been massively reduced.

Smokers in Britain have abandoned the habit over the past 50 years at such a rate that the number of lung cancer deaths is now half what it otherwise would have been. In 1950, 80 per cent of men and 40 per cent of women smoked. Today, one-third of men and women aged 35 to 60 smoke.

Professor Doll was a smoker himself when he began his first study in the late 1940s into the reasons for the sharp increase in lung cancer deaths over the previous 30 years. When the early results started coming in, he threw awayhis cigarettes and has not smoked since.

"Our 1950 study showed that smoking was a cause of most lung cancer in Britain at that time," he said. "Nowadays among people aged over 50 there are twice as many ex-smokers as current smokers and tobacco deaths are decreasing rapidly."

The growing number of ex-smokers set researchers a new challenge - to assess the long-term effects of earlier smoking on people who have given up. The most important finding in Professor Doll's new study, to be published in the BMJ this weekend, is just how beneficial stopping smoking can be. It shows that from the moment a smoker gives up, the risk of lung cancer starts to fall - and if he or she gives up before the age of 35, it returns almost to that of someone who has never smoked. It is never too late to quit.

For lifelong smokers, 16 per cent of men who continue to smoke until they are 75 will die of lung cancer (if they do not die of something else first - smoking increases the risk of many other conditions such as heart disease). The risk of lung cancer falls to 6 per cent for men who give up by the age of 50 and 2 per cent for men who stop at 30. In other words, people who start smoking as teenagers and give up after 10 or 15 years have only a marginally higher risk of developing lung cancer than those who have never smoked.

Professor Doll is living proof of his own research findings. He took up smoking at the age of 18 and continued for 19 years until he was 37, when his research revealed the damage he was doing. Today, 50 years later, he is still working, doing research, publishing papers and presenting his findings to professional and lay audiences.

No one was more surprised by those early results than he was. When he and his senior colleague, Professor Austin Bradford Hill, began their study of lung cancer deaths in 1948, Dr Doll, as he was then, never suspected cigarettes. He believed the cancer was a byproduct of the new obsession with motor cars.

"I thought it was something to do with the road-building programme. We knew there were powerful carcinogens in tar and we didn't know they were in tobacco.

"We looked at occupational studies in road workers but they didn't have higher rates of lung cancer. We were put off [tobacco] by researchers who had tried to produce cancer in animals by painting tobacco derivatives on their skin but the results proved negative. What no one realised then was that tobacco contained weak carcinogens which required exposure over a long period of time to have an effect."

Impressive as the fall in smoking, and in lung cancer, has been over the past 50 years, it is not, as Professor Doll is quick to acknowledge, time yet to celebrate. Despite the gains, smoking is still the biggest cause of early death in Britain, claiming 35,000 deaths from lung cancer in 1998. Half ofthe 12 million people who currently smoke will be killed by their habit if they do not give up.

Women took up smoking later than men and smoked less intensively so their lung cancer death rate has never risen to as high a peak as for men. But the evidence shows that if they continue to smoke they will suffer the same risks. "If women smoke like menthey will die like men," Professor Doll said.

Professor Sir Richard Peto, co-author of the new study and a distinguished tobacco researcher in his own right, said the only way of preventing millions of deaths over the next 50 years was by helping adults give up. Stopping adolescents taking up the habit would save hundreds of millions of lives worldwide but not until the second half of the century.

Professor Peto said: "Tobacco deaths over the next 50 years will be affected much more by the number of adults who manage to stop than bythe number of adolescents who start smoking." He added that if smoking continued at current rates around the world, one billion people would die of its effects over the next century- a thousand times more than the 100 million who died in the previous century.