US scientists have developed a way of predicting how likely a person is to live beyond the age of 100.
The breakthrough, described in the journal Science, is based on 150 genetic "signposts" found in exceptionally long-lived people.
The Boston team created a mathematical model, which takes information from these signposts to work out a person's chance of reaching 100.
It is based on the largest study of centenarians in the world.
This is a rare trait - only one in 6,000 people in industrialised countries reaches such a ripe old age. And 90% them are still disability free by the age of 93.
The researchers now think they have cracked the genetic secret of this longevity.
The team originally embarked on their study in 1995. Since then, they have scanned the genomes of 1,000 centenarians.
They identified genetic markers that are "most different" between centenarians and randomly selected individuals.
The research was led by Paola Sebastiani, a professor of biostatistics at Boston University, and Thomas Perls, associate professor of medicine, also at Boston University.
"We tested our model in an independent set of centenarians and achieved an accuracy of 77%," explained Professor Sebastiani.
"So out of 100 centenarians we could correctly predict the outcome of 77."
She said that the "23% error rate" indicated that, although "genetics is fundamental in exceptional longevity it's not the only thing".
"So there may be other factors like environment or other lifestyles that may help people live longer and healthier lives."
Professor Perls explained that a previous study had looked at longevity in a group of people belonging to the Seventh-day Adventist Church.
"Those individuals have probably among the highest average life expectancy that we know of in the US of 88 years," he said.
"They get there by virtue of the fact that they have a religion that asks them to be vegetarian, they regularly exercise, they don't drink alcohol, they tend to manage their stress well through religion and time with family and they don't smoke.
"To live the additional 10-15 years beyond the age of 88, our paper is indicating that genetics are playing an increasingly important role."
The scientists said that, when it came to genes associated with a predisposition to age-related diseases, centenarians and non-centenarians did not really differ.
"This is very surprising," said Professor Sebastiani. "It suggests that what makes these people live a very long life is not a lack of genetic predisposition to diseases, but rather an enrichment of longevity."
Professor Perls said it was feasible that a simple test could be developed to screen people's chances of being so long-lived.
"I think that that's a possibility down the road," he said. "It brings up this whole field of personal medicine and being able to use genetic information in the future to help guide therapy."
But he added that there should be "a great deal of caution in thinking about what people might actually do with the information".
"Will that stop companies from going ahead and [developing some kind of chip-based test]? Probably not," he said,
"But we think it's really important to understand what people end up doing with this information, including thinking about social entitlements - that merits a lot more discussion."
Professor Sebastiani added: "We have a long list of things to do here.... to understand the real biology behind what we have found."
One of the co-authors of the Science paper is already building a free-to use website where people will be able to use the mathematical model.
On that site, which could be up and running within a week, people who know their genetic code could work out their predisposition to exceptional longevity.
"The site would provide some description of how to interpret the results in the right context," Professor Sebastiani explained.
Dr Jeffrey Barrett, a geneticist from the Wellcome Trust Sanger Institute in Cambridge, UK, cautioned that "subtle biases could make the test seem more accurate than it really is".
"Some of the genetic variants in this study are claimed to have much, much stronger effects on longevity than we've seen in similar studies of diabetes, heart disease and cancer," he told BBC News.
"Evaluation of the test by an independent laboratory will be the ultimate test of its accuracy."
Professor Perls summed up the findings as "a very optimistic message".
"Exceptional longevity is not this vacuous entity that no one can figure out," he said.
"I think we've made quite some inroads here in terms of demonstrating a pretty important genetic component to this wonderful trait, and this really opens the door to future research."