Japanese scientists have used genetic modification technology to create a decaffeinated coffee plant - the first time it has been done, according to expert observers.
By Richard Black
BBC science correspondent
Writing in the British journal Nature, the researchers say their GM plants contain roughly a third of the caffeine content of natural varieties.
The GM coffee plants are about a year old
But it may be many years before engineered coffee beans are being used widely to make refreshing drinks in the world's cafes.
The active lobby against GM technology that exists in some countries has particular concerns about coffee, a crop that is hugely important to developing countries.
ActionAid has described GM technology as "a trend towards corporate control".
It fears novel coffee crops may reduce considerably the job opportunities presently offered in the labour-intensive industry.
Currently, decaffeinated coffee is made by growing normal coffee beans and then processing them to extract the caffeine. It works - but many drinkers believe it removes the taste as well.
The Japanese researchers, from the Nara Institute of Science and Technology, believe their GM plants will provide decaf which actually tastes like coffee.
A team of scientists, led by Shinjiro Ogita, used a technique called RNA interference to reduce the activity of key genes in the coffee plant called Coffea canephora.
Dr Ogita told BBC News Online: "[Our] GM coffee plants are still young. They are one year old and about 15-30 centimetres in height, and have not produced beans yet. It will take 3-5 years to harvest. But they are now growing day by day."
With the global market in decaf worth an estimated $7bn a year, there appears to be commercial potential.
But according to the trade body, the International Coffee Organisation (ICO), several major questions remain - not least over how the cost of the GM technology will compare with the industrial decaffeination process used currently.
Pablo Dubois, from the ICO, said: "The global market for coffee is about $70bn a year, and decaf makes up about 10% of that.
"The main issue would be the economics - is it going to be cheaper to buy GM seeds and grow them than to decaffeinate coffee as we do now?"
Other researchers, some in commercial companies, have been working on GM decaf for several years.
But Professor Alan Crozier, from the biochemistry department in Glasgow University, UK, says this is the first credible evidence that the GM decaffeination concept actually works.
"They've got the caffeine level down by about 70%. They'd need to get in down by 90%, but they've used a clever technique and I think that extra reduction should be possible," he told BBC News Online.
Certain anti-GM groups have promised to fight a bitter battle to stop the introduction of novel coffees in particular.
ActionAid has already led strong protests against the introduction of genetically engineered coffee berries that all ripen at the same time.
This would make harvesting on large plantations less labour intensive and put at risk the livelihoods of millions of poor people, the charity claims.