(Updated 2/19/11 for clarity, rewrite of helmet section)


Please note: after reviewing my references, I have rewritten my opinion on helmets somewhat. Last update 2/20/11.

Helmets are quite valuable for children, and I think also for off-road use. Children invariably fall down, and off roading involves a certain number of rocks and trees. Helmets have a lot to offer against falling, and against rocks and trees. (I’ll cite a study on children:


But how much do bicycle helmets offer adults against traffic? Here is a recent tragedy in which the cyclist was wearing a helmet:


Clearly not every accident is survivable, no matter the helmet.

Ken Kifer was formative on my earlier opinion: http://www.kenkifer.com/bikepages/advocacy/mhls.htm. For one thing, he points out that helmet testing specifications are insufficient for protection in traffic. (Ken’s thesis is against mandatory helmet laws.)

Scientific studies of adult use in traffic provide two conflicting pieces of information: 1) that helmets do reduce head injuries, 2) that helmets don’t significantly reduce mortality in populations.


I think what needs to be understood is that a helmet protects only one vital organ, that it does nothing against the crushing weight of a vehicle, and that even a helmet cannot protect your head sufficiently in many collisions.

In short, a helmet may help but it does not make you safe.

What I want to argue is that discussions of bicycle safety are wrong to focus intensely on helmets. We know that bicycle helmets are deficient for use in traffic. Our focus should be on accident prevention. An ounce of prevention is worth several pounds of the helmet cure.

All that being said, I personally wear a helmet because it doesn’t hurt, and anything that improves my chances of reducing injury or escaping death is worth it, so long as the prevention is reasonable and the chances are significant. I think a helmet offers significant protection, if incomplete. (I also need something on my head to tie my flashing P7 LED headlight to, as a measure of prevention.)


Technique belongs right at the top of a bicycle safety discussion. Your most valuable tool of prevention is your brain. There are many good web sites on this subject, but here is how I ride, and a few pointers:

  • I obey traffic laws and behave like traffic while riding on the road. In order for traffic to move expeditiously and safely, it is necessary that everyone obey the same rules and behave predictably.
  • Cross traffic is dangerous - motorists tend not to see bicyclists. This danger increases with cycling speed - it isn’t as big of a deal for children riding at 5-10 mph, but it is a critical issue for adults at 15-25 mph. Behaving unpredictably, riding against traffic or on sidewalks, and wearing poor choices of clothing can make it impossible to be seen. But even if you do everything right, you can still be overlooked by a careless motorist. Any improvement in visibility is worth it. Moreover, it’s my job to second-guess motorists when I have the right-of-way. I can slow down and be prepared to stop. I also point my helmet light right at a car to improve my chances.
  • I never ride on a sidewalk unless I need to use a roadway that is too dangerous for me. When I enter the sidewalk, I go into “pedestrian mode,” which is around 5 mph (very slow). It’s much better to find an alternate route, but occasionally there isn’t one.
  • Visibility in daylight - bright clothing is the key. Yellow, neon orange and neon green are attention-grabbing. Bright orange and bright green (not neon) are okay. White and red are not attention grabbing, and darker colors are invisible. I would also like to find a helmet with a mirror-finish. A round object (such as a helmet) with a mirror-finish has a fantastic tendency to reflect sunlight in all directions, providing an attention-grabbing glare, although that only works in full sun.
  • Visibility in dark, dusk or overcast - a headlight, helmet light, taillight, and reflectors are crucial equipment. Bright enough headlights may also help a little in daylight - P7 LEDs (to a lesser extent Q5 LEDs) are about on par with other vehicle headlights. You may notice regular vehicle headlights help a lot after dark and on overcast days, but only help a little bit in broad full sunlight.
  • (Note that the “reflectors” that you wear are actually “retro-reflectors” - when a headlight shines on them they throw the light back on the origin. Unless conditions are dark relative to vehicle headlights (dusk, dark or very overcast), reflectors do nothing, unless they are also yellow or neon, which is only as attention grabbing as proper clothing. http://en.wikipedia.org/wiki/Retroreflector )
  • The big judgment call is whether to take a lane or ride on the shoulder. I spend most of my time on the shoulder, but it’s based on my judgment that the shoulder is safe and wide enough for cars to pass. Generally I take the lane through certain intersections, narrow roads, and areas with parked cars. Moving off the shoulder and into the lane is the same as a lane change - it requires a check over the shoulder. In one segment on my route, the road is four lanes with no shoulders, and so I always take the whole right-hand lane. It’s possible for this cause problems on four-lane roads if the traffic is heavy, where tailgating drivers may attempt a pass in the right lane (my lane) without seeing me, but the traffic is generally pretty light and I’ve judged this to be safe. Cyclists need to be assertive when taking the lane, but also be wary of traffic approaching from the rear.
  • My rear-view mirror of choice is a Take-A-Look eyeglass mounted mirror. Helmet mounted mirrors are also great (maybe best). Not all bar-mounted mirrors are created equal, many of them sit too far out of view and/or have too much curvature to be effective as a rear-view, but some of them are quite good. The purpose of a mirror is to increase situational awareness. As I approach crossing traffic while I have a right-of-way, I know that I am not in danger if I’m simultaneously being overtaken by another car. However if there will be a break in my pack when I pass the crossing traffic, or if I’m alone, that is a reason to heighten suspicion and be looking for an out, or to slow down. Many cyclists ride without a mirror, and I consider it non-essential, because you can also use your ears and head checks to be aware of your situation. I prefer a mirror because it reduces the number of head checks, thus keeping my eyes in front of me for a greater proportion of the time, and I ride in traffic that is heavy enough that I can’t always pick individual cars out of the noise. Not all head checks are eliminated by a mirror - I still usually check before I change lanes (or when I come off the shoulder and move into a lane, which is the same thing). A rear-view mirror may also help prevent being hit from behind, but then it might not. Fortunately being hit from behind is fairly uncommon, the motorist has to be quite incompetent for it to happen. (But don’t get me wrong, there are drivers out there who will do it, so it’s best to assume the worst if you see someone coming who makes you uncomfortable.)
  • One newbie mistake is to swerve into traffic when going around an obstacle. I treat this as merging into the lane, and if for some reason I can’t enter the lane due to traffic, or the obstacle came up suddenly and I don’t have time to check over my shoulder, I may slow or come to a stop. Just the other day I came up on a parked car on a narrow crowded roadway, but I saw it ahead of time and signaled and moved into the crowded lane. I had enough space to signal twice and the car (finally) reacted and let me in. Speed is also a factor on a crowded roadway: in this case it helped that I was going over 20, the car was only going about 25mph. If the cars had been bumper-to-bumper at 35 or if I had slowed to 15 then my maneuver wouldn’t have been possible that day, I probably would have had to slow and creep around on the sidewalk.
  • There are several situations that are difficult to handle, even with plenty of experience. One is a crowded uphill roadway that is too narrow or has parked cars. If I were to take the lane I would be a major obstruction. This situation can provoke drivers to attempt dangerous passes inside the lane. If available, I will go to the sidewalk, since I’m slow on the hill anyways, and use my pedestrian mode. If the traffic happens to be gated by a nearby light and the hill is short enough, I may stop and wait for a red light and try to clear the hill between lights (although some lights are crowded in multiple directions and the red light doesn’t effectively “gate” uphill traffic). Fortunately I’ve been able to plan my regular routes to avoid this situation.
  • Another hairy situation, heavy bumper-to-bumper traffic, is not too bad if the speed is low enough for me to keep up at a similar speed, but not if the traffic is at high speeds. Even at low speeds, paranoid attention is needed in to avoid cars suddenly turning.
  • Note that bells work with pedestrians, but are not for traffic. I have a Delta Airzound, which is a loud push-button air-horn with a compressed air reservoir, although I seldom use it. Many cyclists don’t use any horn at all, since hands are needed for brakes in a crisis. Shouting (a.k.a. screaming) is a nice hands-free and very audible way to grab attention in an emergency. The real purpose of the Airzound is to educate drivers. (“Hey buddy, don’t buzz me next time, all-right?”) Also, if someone honks at me, I honk back, presumably they were trying to make friends.
  • One resource that I’ve found invaluable is Google’s maps, with satellite photos. Most of my miles are on regular routes, so I’ve done a fair amount of research inspecting and choosing routes that have the best margins, the least traffic, and the fewest stops.
  • A cyclist’s skill level is quite important. If you just recently acquired a new bike or are just starting out, it’s best to break it in where there is no traffic. Don’t go into traffic until you are comfortable with your equipment.
  • Those more familiar with cycling technique will note that my above notes outline “vehicular cycling,” but this is not the only school of thought. For example, there are a few cyclist who really believe that riding against traffic or on the sidewalk is better. I disagree overall, however I don’t think that is so fundamentally unsafe or slow that it can’t be done - but anyone who chooses to ride against traffic needs to really understand that they are generally invisible to everyone. If you have the right mindset, maybe that’s okay, but not for me.



Safety Statistics

Statistically speaking, there are about 800 bicyclist deaths every year in the US. It would be helpful if I could somehow calculate my own risk, using a comparable population riding a comparable number of miles on comparable types of streets. Too bad that is impractical. In fact, I can’t even get a hard number on the basis of risk per mile, because we don’t know with much accuracy how many miles US bikers ride per year. Instead of trying to work this out myself, I’ll refer to Ken Kifer:


Ken used the high number 21 billion miles per year, essentially because he couldn’t believe the low number 6 billion. According to his calculation, bicycling is around 2.5 times more dangerous per mile (in risk of death) than driving or riding a car, which makes it about the same risk per hour. If by chance Ken is wrong and the low number is correct, then the risk could be another three times greater than that.

(Ironically, Ken Kifer was killed by a drunk driver in 2003. It doesn’t make me feel a lot better to know that the driver was convicted of murder for this. I hope that such cases would act as deterrents to drivers who would endanger others, however I’ve seen a lot of bad driving, and I don’t believe it. If I were killed by some jerk, he indeed ought to be prosecuted as a criminal, but it really wouldn’t make that much difference to me (being dead), and I doubt that it made much difference to poor Ken.)

Assuming that I ride my bike to work 3 out of 5 days from now until I’m 60, I would ride a total of almost 100,000 miles, divide by 21 billion and multiply by 800 and my statistical lifetime risk of death on my bike commute appears to be about 0.4% or 1:270. On the other hand, if I use 6 billion instead of 21 billion, I get 1.3% or 1:76 lifetime risk of death while commuting. Either number doesn’t appear to make a huge increase in my risk of traffic mortality, compared to the average American’s lifetime risk of dying in traffic (around 1:60 or 1.7%).

(Part of the reason this is so marginal is that I’ve put my annual bicycling miles at only about 3,000 per year, whereas the average American drives over 10,000 miles per year, also I’m limiting the age range over which I’d be doing this (more on that below). The risk per mile is higher, but the total number of miles is much much lower than that I will drive.)

The problem with statistics is that they apply to populations, not to individuals. Real world risk is not per mile, but rather risk is per pass, per intersection, and per lapse in judgment, and it depends a great deal on how I ride, which roads I choose, and how vigilant I am. (Of course there is a completely random element as well. That drunk loser is going to kill someone tomorrow night, if I’m in the wrong place at the wrong time I probably can’t prevent it from being me.)

I suspect driving habits in the car are an indicator of bicycle safety. I’m not as safe of a driver as I could be, but I think I’m reasonably safe - I’ve had a few very minor collisions, but I like to say that I’ve never bent a fender. (As an aside, I’m totally mad at my mistakes and I swear I’m going to do better. It’s not as easy as it sounds: traffic habits and personality are intertwined.)

A few sources have pointed out that the health benefits of exercise outweigh the risks of cycling in traffic by 6:1 or up to 20:1. However, there are certainly much safer ways to get exercise, such as a treadmill or a stationary bike. I think most people agree that walking or jogging (on the street) is not safer than bicycling, although I believe that conclusion depends a great deal on the choice of roads, time of day, dress, and it may not be true at all when you look at it per Calorie instead of per mile.

In my own case, I own stationary exercise equipment, and my employer gives out almost free gym memberships, but I still feel that there are benefits to exercising while commuting - because I have to commute on a very regular basis, but I don’t have any particular deadline for exercise - it makes it easier for me to make time. I also gain a little time that I would have spent driving. Also, let’s face it, stationary aerobic exercises are intolerably dull. Anyone who has actually managed to stick to a routine on a stationary bike or treadmill - good job, I can’t do that, you’re a wacko.

Lastly, it has to be pointed out that failure to exercise creates a risk that crops up starting around mid-life, generally after age 45 and especially over age 55, whereas cycling now creates a risk right now. If my big worry was dying before age 40, then not exercising is the way to go. (Or exercising on stationary equipment is even better.)

It’s also worth pointing out that accident survivability drops with increasing age. If cycling is really for me, it’s still worth considering that maybe I should get off the bike and move to a safer exercise when I get to around age 55 or 60.

Get off the bike just when I reach heart attack age? Where are the health benefits then? First of all, I think there’s a big difference between reaching 60 in excellent fitness versus reaching 60 in decrepit fitness. It’s difficult for unfit people to become fit, and this difficulty becomes steeper and steeper with age, as joints and bones grow weak and uncooperative, muscle recovery grows longer and more painful, and habits become mentally cemented. Second of all, I’m expecting (or trying) to achieve some other lifestyle changes that will fit into this whole scheme. Namely, that I want to be self-employed (and not commuting) before around 45 or 50, and especially I want to be working fewer hours before 55 or 65, such that I might have a lot more recreational time in which I can have more fun exercise, such as hiking and swimming.