As the world's top athletes prepare for the Sydney Olympics, Geoff Watts asks if we can continue to set new records or if our species is fast reaching its best.

Since 1954, when Roger Bannister ran his mile in 3 minutes 59.4 seconds, the world record for that mesmerising distance has been broken 18 times. Five seconds had already gone by the end of the 1950s. Three more were lopped off in the 1960s, another three in the 1970s, but just two in the 1980s.

The 100m shows a similar pattern. Between 1930 and 1960, 0.3 of a second was shaved off the world record. Between 1968 and 1999 (the authorities having sensibly begun measuring in hundredths of a second), it fell by only 0.16.

This law of diminishing athletic returns applies to most sports in which records are a direct reflection of human performance. We are likely to see it again next month at the Sydney Olympics. But will the diminution eventually fall to zero, meaning that the age of record breaking is finally over?

ÖйúAƬ

If we humans were machines designed on drawing boards, constructed of metal, plastic and ceramics, manufactured to precise tolerances, subject to wear but not to repair or improvement, powered by a single type of fuel and controlled by an inflexibly programmed computer, it would be possible to calculate the precise limits of our performance. But we are not.

We are not designed. We grow individually, each according to a unique set of inherited instructions and subject to all sorts of outside influences. We are made of protein and carbohydrate combined in varying proportions. Much of our wear can be repaired. Our fuel is bewilderingly various. And the entire system is controlled by a brain in which states of mind have dramatic effects on physical performance. In short, like all biological machines, we are diverse. No human will ever lift a five-ton weight, but where the limit lies is a matter of conjecture.

ÖйúAƬ

Most academics seem to think there is still a way to go. Peter Radford, sometime runner and chair of the sports science department at Brunel University, says we are still recruiting new populations into sport who will produce new talent to force the limits.

And there will always be exceptional individuals. Tom Yule - a weightlifter studying for a PhD on biomechanics at Manchester Metropolitan University - says that records have been static in recent years. "But there is an 18-year-old Pole who has already won the European senior championship. He is lifting weights that people who have been at it for years cannot manage."

More scientific training methods may yet squeeze extra performance out of athletes. Nutrition has contributed - generally, in that our entire population is bigger and stronger, and specifically, as part of every training regime. Clyde Williams, professor of sports science at Loughborough University, believes there are still gains to be made in using nutrition to help the body recover from exercise. A slow rate of recovery may limit the frequency with which an athlete can exercise and thus the attainment of peak performance. Williams is trying to find nutritional methods of speeding the recovery process.

Craig Sharp, a cross-country runner and another Brunel sports scientist, speaks of "micro-nutrition" - the attempt to move from the steak-and-cereals level of categorising food towards biochemical breakdowns.

Could this really make a difference?

"Yes, at the higher end of sport where they are looking for small increments in performance," Sharp says. "The difference between a gold medal and not getting to the final in the 100m is about 1 per cent."

Another issue being tackled is the build-up of lactic acid within hard-pressed muscles. "People devise training regimes that produce more power, but also more lactic acid," Sharp says. "You have to follow regimes that maximise lactic acid removal. Ten years ago, nobody would have bothered."

Athletes may try to neutralise some of the acid by taking sodium bicarbonate - another small advantage: 0.2 or 0.3 of a second over 400m. But that may be enough.

There is also intense interest in "ergogenic aids": substances such as creatine, which athletes can take to help store energy.

ÖйúAƬ

For all that, Sharp believes that physiology has less to offer than psychology.

Sandy Wilson of Northumbria University, one of a growing number of accredited sports psychologists, agrees this still has scope for helping athletes reduce anxiety and boost concentration. "Some people feel there is a stigma in psychological help," she says. "They have a false idea about what we do and are uneasy about seeking our help."

ÖйúAƬ

Technology is another part of the equation. Chris Boardman's bikes, new surfaces for running tracks, swimming costumes that mimic the flow pattern of water over a shark's skin - these and scores of other ideas are raising speeds and reducing times.

The potential varies from sport to sport. Ron Thomson of the sports engineering group at the University of Glasgow does not think that running shoes - despite the glowing TV adverts - have much more to offer performances. He also criticises so-called "energy return" - a springy effect in which energy stored in the shoe as it is compressed is released as the athlete's toe lifts off.

"When speaking off the record and not sponsored by one of the running shoe companies, most people agree that this is a load of nonsense," he says.

But physicist and pole-vaulter Colin McCarthy swears by the characteristics of the pole in pole vaulting. While he says the material of the poles has come on considerably over the past 20 years, he can still see scope for improvement.

Nutrition, biochemistry, physiology, biomechanics, psychology are all in there, doing their bit to help athletes run, jump and throw faster, higher and farther. But one science is missing - pharmacology. There is no shortage of experimentation, no lack of subjects keen to try their luck. Yet the testing is haphazard, the results poorly measured, and the benefits - even when manifest - are neither published nor celebrated. Sport (officially, at any rate) continues to operate its ban on the exploitation of this branch of science.

Suppose pharmacology were to be granted its place: would we see more records being broken? Calvin Morriss of Manchester Metropolitan University's sports science department - currently seconded to UK Athletics, the governing body - has no doubts. "You certainly would. Look in the medical and pharmacological literature. And we know from people who have been caught taking drugs and turned in outstanding performances that their effects are not just psychological. If you could take whatever you liked, I'm sure there would be improvements."

Will it happen? "Not at present. But there might be a gradual change. People might see that athletes are taking every other aspect of their performance to the nth degree and ask why don't we allow them some amount of freedom in this area too."

Radford also suspects that drug-taking has contributed substantially to existing records. He too expects to see drugs increasingly tolerated. "Everyone seems to take creatine phosphate and a lot of athletes take growth hormone, which they know is not being tested for. The legal profession has been undermining the anti-doping programme over the past ten years. Eventually they will succeed because sport will not be able to afford these disputes."

There is one other factor that could affect the sports community in the future - our knowledge of the human genome. Most elite athletes, however hard they train, would probably accept that some inherited quality, be it leg length, joint structure or lung capacity, has contributed to their success. Armed with their working draft of the human blueprint, scientists will try to identify the role of each of its component genes. They will compare the genes of the healthy with those of the sick to look for differences that might explain disease. But similar comparisons could be carried out on genetic material from elite athletes. Considering the profit and national prestige bound up with sport, the expense of such an exercise is unlikely to be a barrier.

If you could identify genes that predispose to success in sport, how would you use them? Forget about breeding super-athletes: too slow and too unreliable. What you might do is what doctors will do when exploiting genetic discoveries in medicine - mimic the useful genes by administering something that replicates the actions of whatever product they code for; or suppress the action of unhelpful genes by inhibiting their products.

But many of the chemical agents required to perform these tricks would be categorised as drugs and therefore banned.

Meanwhile, the evidence suggests that records will continue to be set, but with the graph line of performances curving evermore asymptotically as it strives to reach the plateau of what can be achieved.

ÖйúAƬ