'Why on earth," says Anne Portehault, chaining her bicycle to a railing opposite the Hôtel de Ville in Paris, "would I want to put a cycle helmet over my hair? I would look ridiculous." Well, not ridiculous, I counter, but you would certainly look less chic and enviably coiffed than currently. The 39-year-old is angry enough that from the autumn, Parisian cyclists will be required to wear fluorescent jackets. "They only cost a couple of euros but they make anyone who dresses well - such as me - look absurd. That yucky yellow - I've got nothing to match it in my wardrobe. What am I going to do?"

But what about the danger of your brains spilling out of your head after being knocked over by one of the French capital's aggressive drivers? Isn't ridiculousness a small price to pay? I remind Portehault that it was nearly on this spot, outside the BHV department store on the Rue de Rivoli, in 1980 that the life of one of France's great penseurs, Roland Barthes, was brought to an end when he was run over by a van. "I pay a lot of attention to the traffic," replies Portehault. "I'm very careful when I cycle. So, no, I don't think looking ridiculous is worth it."

It's worth pointing out that it's not only Parisian women who worry about their hairdos being mangled by helmets. A young man chaining his bicycle to a lamppost, who declines to give me his name, says: "I've spent ages gelling my hair. I'm not going to ruin all that work, thanks very much." I find Marco Simioni, 45, a Paris-based architect and artist who originally hails from Vicenza, poised to ride off on a hired cycle. "I've only been cycling for four months and I try to be very careful, but it's true I run a risk not using a helmet. It's a risk I'm prepared to take."

I've come to Paris to find out why there is such a difference between French and British cyclists when it comes to wearing helmets. A few years ago, when I worked in Paris for the Guardian, I always wore a helmet when I cycled through the Marais to my office near the Opéra Garnier and looked in disdain at the flowing locks of my more hirsute fellow cyclists as they cheated death and looked glamorous. Now, thanks to mayor Bertrand Delanoë's bicycle rental scheme, Vélib, which was introduced just over a year ago, there are more Parisian locks flying in the wind than ever.

In France there is a slang phrase "C'est relou", which translates as "It's boring." "C'est relou de porter un casque," several cyclists told me. It's boring to wear a cycle helmet. But not just boring: ugly, unstylish, fundamentally un-French.

"There is a great cultural difference between the English and the French," says Anne Portehault. "We are very, very independent and very rebellious. This is the home of liberty, after all." Portehault, who rides horses in the morning and works in an American law office in the afternoon, doesn't use the sturdy Vélib bikes that can be hired from ranks one finds every 300m in Paris. Instead, she favours a venerable yellow number dating from the 1960s, which used to belong to her sister. "It works but it looks rubbish. So no one's going to steal it."

I'm not sure that Gallic rebelliousness tells the whole story of why, if you cross the Channel, helmet-wearing cyclists become much rarer. Perhaps, and this is just a theory, the British are more risk-averse than continentals. Indeed, the difference is not just between Britons and continental Europeans, but between many anglophone countries and others. Cycle helmets are common in the UK, the US, New Zealand and Australia. But if you go to some of the cities where cycling has a prouder history than in Britain - cities such as Amsterdam, Copenhagen, Berlin or Shanghai - you will find very few cycle helmets.

Of course, it's unfair to over-generalise. On a nearby side street, I find a Parisian cyclist wearing not only a helmet, but a fluorescent jacket and sensible gloves. Why, I ask 35-year-old Laurent Marin, do you not bend your knee to Paris's prevailing sartorial strictures for cyclists? "A few years ago I was cycling at the Porte de la Villette when I saw a motorcyclist was run over by a car. It was horrible. I rang my wife and said: 'I'm going to be late because I'm going to buy a helmet.' Now I won't let my wife or my children go cycling without helmets." Courier Felix Nizard, 20, tells me that he wears a cycle helmet his friend gave him. It seems several sizes too big, I say. "Yes, but it looks kind of cool with the advertising all over it," he retorts. "I feel happier wearing it now with all the Vélibbers around. They are making the city more dangerous."

The number of cycle journeys has grown exponentially in Paris since the introduction of the Vélib scheme, and many of the inexperienced users do indeed "cycle like idiots", as Marin says. That said, there have only been three reported cases of Vélibbers being killed. On France's bicycling blogs, this is a hot issue. One blogger, a man who styles himself Ludovic Bu, wrote recently: "The figures are clear: three deaths in a year and 24m journeys on Vélib bicycles. That amounts to a 0.0000125% probability of killing yourself while using a Vélib bicycle." But he adds that in Paris last year, 53% of those who died or were injured on Paris streets were motorcyclists, 22% pedestrians, drivers of four-wheeled vehicles 18% and cyclists only 7%. "Thus, one can well see that it is less dangerous to travel on a bicycle than on a motorbike, on foot or by car."

I'm not at all sure these figures warrant the inference that cycling in Paris is safe, but that doesn't matter: such rhetoric contributes to a blithe attitude on the part of Parisian cyclists towards sustaining death or serious injury. It makes them less likely to even consider whether wearing a helmet would be worthwhile.

Delanöe's aim, according to many Parisians I spoke to, is to make Paris more like Amsterdam, where cycling is such a venerable and popular mode of transport that cyclists ride without helmets with a confidence that actually has some foundation. One Dutch idea that both the British and French authorities might consider introducing is the law of "strict liability" whereby, in any collision between a cycle and a car, the motorist is deemed to be responsible. This makes motorists much more wary. And what about the woonerf, a street or group of streets where pedestrians and cyclists have legal priority over motorists, and where motorised traffic is restricted to a speed limit of "walking pace"? In Germany, they have similar zones called Verkehrsberuhigter Bereich, where motorised speed limits are restricted to 7km/h. If only Paris's Rue de Rivoli or, say, London's Embankment became woonerven or Verkehrsberuhigter Bereich tomorrow, then I might consider leaving my helmet at home. But they aren't, so I won't.

Matt Seaton, the Guardian's former cycling columnist and author of On Your Bike!: The Complete Guide to Cycling, distinguishes sharply between cycling in continental northern Europe and elsewhere. "In northern Europe, especially in countries such as Holland and Denmark, there is much more of a consensus that bicycles have priority. In Britain, it's much more of a war of all against all. France is somewhere in between. In Britain our road environment is much more of a free market and that may make cyclists ride more quickly and aggressively than they do in the rest of Europe."

It is also worth pointing out that in Britain we tend not to ride the sit-up-and-beg bicycles common in Holland; we ride mountain bikes and racing cycles that, you might well think, facilitate aggressive cycling. I once saw a cyclist respond to being abused by pedestrians for jumping a zebra crossing by leaping off his bike waving a D-lock like a mace. There is a medieval flavour to London's cycle scene. In that road environment, British cyclists dress up like extras from Rollerball or Blade Runner, almost as if they are going to war and expecting serious injury. The cycle helmet is part of the uniform.

But is wearing a helmet actually going to save one from injury? This is a vexed issue among cyclists. The fact that London's tousle-haired Tory mayor Boris Johnson now eschews a cycle helmet has focused the debate. "It's a very bad example for the leader of a capital city to set," said Simon Turner, a 41-year-old cyclist who suffered permanent brain injuries as a result of falling from his bicycle in Richmond Park. "He's saying, 'Don't worry, take the risk.'"

"In my efforts to do the right thing," Johnson replied in a newspaper column, "I have ended up giving offence to both opposing factions. As soon as I started to wear a helmet , I was denounced as a wimp. As soon as I was pictured not wearing a helmet, I was attacked for 'sending out the wrong signal'." Cyclists, he argued, should have the right to choose "hatless, sun-blessed, windswept liberty" or " helmeted security".

What would happen if Boris's brains were spilled, say, at the cycle lights next to Islington Central Library on Holloway Road near the mayor's home? Would cyclists pull up to his twitching, prone form, solemnly remove their helmets and say, "Told you so"? The issue was recently aired in the London Evening Standard, where columnist Dr Mark Porter defended Johnson's decision to cycle helmetless on the grounds that in over 20 years in medicine he had never had to treat a cyclist with a life-threatening head injury. "Indeed, the Bicycle Helmet Research Foundation estimates that the average cyclist would have to pedal the roads for more than 3,000 years to suffer a serious head injury, let alone one that would be mitigated by a cycle helmet," wrote Porter. "Research suggests that the mere donning of protective gear such as helmets encourages cyclists to take more risks on the road, and therefore, paradoxically, increase their risk of serious injury." Porter added that he was happy for his children to cycle without helmets.

This column provoked a letter from Dr Andrew Curran, consultant paediatric neurologist at Alder Hey hospital. "The probability of having an accident resulting in a serious head injury which could be mitigated by a helmet is mercifully low: there are many kinds of high-impact crashes where a helmet, or any form of body protection, would not save you. But the issue is that if an unhelmeted child is involved in a non-fatal crash resulting in a serious head injury, there is a 50 to 70% chance it will result in a serious cognitive deficit . . . It seems sensible that, just as we immunise children against serious diseases they are very unlikely to catch, we ensure they take precautions against the possibility of head injuries, however rare. The blase attitude of Boris Johnson and cycle campaigners on this point surprises me."

The rise in cycle helmet use in the past quarter of a century suggests that Johnson is swimming against the tide. From the mid-80s onwards, a steadily increasing proportion of cyclists has worn helmets. The Department for Transport found in 2006 that cycle helmet-wearing had risen from 28.2% on major roads in 2004 to 30.7% in 2006, while on minor roads the figures were 9.6% and 13.8%.

The total number of deaths of cyclists in the UK has been falling almost steadily since 1934, when there were 1,536 cyclist deaths. In 2007, there were 136. No one, however, has yet demonstrated that the rise in helmet use has made any contribution to that decline. Indeed, there is a statistic that associates a rise in helmet use with an increased risk of cyclists suffering

serious head injuries. When cycle helmets were made compulsory in Australia the number of people cycling fell by a third, and although the number of serious head injuries fell, too, it only fell by 11%. So, despite a large rise in Australia in helmet use, the risk of head injury among those who continued to cycle actually rose. Of course, one can't infer from this that cycle helmets cause serious head injuries, but the statistical association of a rise in both helmet use and in head injuries is embarrassing for those who advocate helmets.

What is a bicycle helmet supposed to do? The short answer is that it is supposed to reduce the risk of both skull and, more importantly, brain injuries caused by the brain jolting around inside the skull as the result of collision. According to the DfT's paper Bicycle Helmets: A Review of Effectiveness, "While simple fractures to the skull can heal, brain injuries, unlike those to other body regions, do not and can lead to long-term consequences. "A bicycle helmet is supposed to perform three functions, namely: "reduce the deceleration of the skull; spread the area over which the forces of impact apply; and prevent direct impact between the skull and impacting object."

The DfT paper concludes: "There is now a considerable amount of scientific evidence that bicycle helmets have been found to be effective at reducing head, brain and upper facial injury in bicyclists."

How are they supposed to do that? According to the DfT: "These three functions can be achieved by combining the properties of the soft, crushable material [usually polystyrene] that is incorporated into helmets, usually referred to as the liner . . . and the outer surface of the helmet, usually called the shell." Historically, helmets had hard shells but now the tendency is for there to be no shell at all or a very thin shell.

The blogosphere, however, seethes with sceptics who argue that bicycle helmets do not work when they are struck by fast-moving vehicles or when a cycle helmet hits a kerb or road at speed. Some argue that they may produce a false sense of security and a cavalier attitude to risk. "Their design-intended strength is equivalent to a impact speed of about 12.5mph," writes cycling blogger Patrick Herring. "They were never intended for collisions with cars. They're not a cycle-equivalent of motorbike crash helmets (and you can't wear one of those because your brain would boil). They work by the outer shell keeping the polystyrene in place whilst it absorbs the deceleration by being crushed. Counter-intuitively, if the shell breaks in the initial contact the total energy absorbed is a lot less: a broken helmet is one that didn't work."

This, he says, means that above about 12.5mph the helmet has little effect; certainly it won't make a 30mph crash feel as if it happened at 17.5mph.

One psychological study, meanwhile, suggests that helmets make cyclists more, not less, vulnerable. Two years ago, Dr Ian Walker, a travel psychologist at the University of Bath, had his bicycle fitted with a special sensor and then went off riding, sometimes wearing a helmet, sometimes not, and sometimes wearing a wig intended to make him look like a woman. It is, incidentally, marvellous that a) there is such a thing as a travel psychologist and b) that they get to do such fun-sounding research.

The sensor measured the proximity of passing traffic. When Walker rode bareheaded, cars gave him a wider berth than if he was wearing a helmet. They gave, on average, an extra 85mm. Dr Walker concluded that this was because his dearth of proper kit indicated a lack of cycling competence that might result in wobbling, unannounced 45-degree turns, unscheduled wheelies etc. Helmetless, he looked more vulnerable. When he wore a helmet, motorists passed closer. On two occasions they passed so close that they hit him. "One of the main findings," says Dr Walker, "was that a behaviour intended to reduce risk (putting on a bicycle helmet) might paradoxically increase one's overall level of risk because drivers react to its presence by changing their behaviour."

But it's what Dr Walker did next that I find particularly suggestive. He cycled about wearing a black wig that made him look - at least to his satisfaction - like a woman from the rear. He found that motorists gave him a wider berth than normal, which he ascribed to a gender effect: motorists are more considerate of women rather than men cyclists. Of course, there is another possibility, namely that Dr Walker was so poor a female impersonator that motorists, possibly out of solidarity with real women, or conceivably because they thought he looked barmy, steered clear. I haven't seen the photos of the experiment, so I can't be sure.

What I find interesting is that this research suggests that women cyclists with more hair than is fair (or men cyclists impersonating such women) are less likely to be run over than mugs like me who wear cycle helmets. Which brings us back to Anne Portehault.

I forgot to tell her that there is a new, putatively stylish cycling helmet cover devised to "encourage females to cycle". The "cycle head dress" is essentially a hat that you can put over your helmet to make yourself look chic. Have a look at their range at hi1-designs.com.

Portehault probably wouldn't have liked the cycle head dress. After all, Parisiennes like her have no need of helmets or helmet covers, and they have research from Bath University to support their lifestyle choices. There she goes now, cycling down the Rue de Rivoli, one of Paris's more terrifying one-way boulevards, wind in her hair, as chic as can be as the photographer takes her picture. You'll have to excuse me for feeling more than a little jealous. If only I had more hair and was a woman, I think as she rides off towards the Palais de Louvre, I might feel safe without a helmet as I cycled. Or maybe I should get a wig.

Have your say: Should cyclists wear helmets? blogs.theguardian.com/news