Is Bicycling Safe? Is Bicycling Dangerous?

Short answer: As with all modes of transportation, it entails some danger.

Longer answer: yes, similar to the risk of motoring — perhaps twice as risky. but how to measure? (per mile, per trip?). Bike-MV collisions are currently running 2% of all in AZ. Bicycling represents perhaps 1%, i.e. twice the risk.

For the moment, this is going to be a catch-all for links and related info on the topic. Links:

Good factoid in the article about a bicylist who plead guilty to a felony (thought to be the ONLY one in history) vehicular homicide in a pedestrian death in 2012 in San Francisco:

Some 4,834 cyclists and 59,925 pedestrians were killed by motor vehicles in the United States between 1999 and 2009 (the most recent year for which figures are available), according to the Centers for Disease Control and Prevention. Cyclists killed just 63 pedestrians, or about six a year, during the same time period. — NYTimes

That’s about a thousand-to-one ratio of car-ped to bike-ped pedestrian deaths. Car drivers are more dangerous. That was re-quoted by Carl Alviani‘s article widely disseminated entitled Why Bikes Make Smart People Say Dumb Things:  “After 15 million miles traveled, the Citibike program has still caused not a single fatality for either pedestrians or riders” (stats as of 2014?) he also ; which lays out a good case that bicycling isn’t particularly dangerous to either riders or other road users. The bikeshare stats were sourced to a slate article Not One Person Has Died on an NYC Bike-Share Bike;

Many places have (re)cited data attributed to a company named Failure Analysis Associates, Inc  (now know as Exponent), expressed as # of Fatalities per 1,000,000 Exposure Hours

  • Skydiving 128.71          Snowmobiling .88
  • General Flying 15.58      Motoring .47
  • Motorcycling 8.80      Water skiing .28
  • Scuba Diving 1.98         Bicycling .26
  • Living 1.53                    Airline Flying .15
  • Swimming 1.07           Hunting .08

I have been unable to find any primary sources on this data and how it was produced. Though a relatively recent Grist Magazine article Not pedaling can kill you, Alan Durning  states “The engineering journal Design News published it with little comment in 1993 in an article on a different subject”, and has some futher elaboration.

Take Back the ‘Burbs, Sunset Magazine June 2012 issue;  sidebar titled Is Biking Safe?:

Fear of traffic is one of the top reasons people don’t bike. But, statistically, biking is safer than driving, and wearing a helmet makes it even more so. Here are the annual odds of fatalities for common activities, according to the most recent stats. –Aislyn Greene

  • Motorcycling 1/ 6,141
  • Driving 1/ 11,883
  • Working 1 / 30,735
  • Biking (all) 1 / 68,673
  • Walking 1/ 75,026
  • Swimming 1/ 87,357
  • Biking with helmet 1/ 342,847
  • Flying 1/ 1,476,136
  • riding in a train 1/ 216,475,677

Sources: Bureau of Transportation Statistics, Centers for Disease Control and Prevention, Consumer Reports, Federal Railroad Administration, Insurance Institute for Highway Safety, Motorcycle Industry Council, National Transportation Safety Board, the Outdoor Foundation, U.S. Bureau of Labor Statistics, U.S. Department of Transportation (full article currently available at site)

Not many clues as to how these numbers were figured.


Even though we’re “guilty” of citing same NHTSA fatality stats in our WP’s without qualification, I spent some time during interview explaining that “exposure not reflected in those stats and I’d recommended he provide that qualification to readers (e.g.,; etc.  I had someone sent me a reminder: 
Chance of death by motor vehicle: 1 in 84;
Chance of death by pedestrian collision: 1 in 626;
Chance of death by bicycle collision: 1 in 4,919 
And then there was the part about taillights that still needs correcting . . .
I’d recommended he provide link to our Arizona Bicycling Streets Smarts ( and STRguide ( And, I was hoping that any quote from the other guy would have focused on skills/knowledge cyclists can learn/gain from an Effective Cycling / Traffic Skills class.
But as we’ve pondered before:  how do we reach the “transportational cyclists” out there right now (and tonight without lights) — crossing all those freeways going the wrong way on the sidewalk!

A “scholarly” reference — “The Invisible Cyclists of Los Angeles,” Progressive Planning: The Magazine of Planners Network, Summer 2010 (on-line at  “Thousands of working-class people use bicycles to traverse cities and towns across the U.S. every day. . . . this group of cyclists is as dedicated as any other, riding through the wet of winter and simmering heat of summer. . . . you won’t see invisible cyclists at . . . City Council meetings demanding bike lanes. You might not see them in the street either, as these cyclists tend to ride alone, often intermingled with pedestrians on the sidewalk, and without lights or reflective clothing. . . . Low-wage workers have limited transportation options, compelling them to bike. Since work may not be steady enough or income high enough to be able to afford a car, or perhaps even a monthly bus pass, some are effectively captive cyclists. Limited mobility means fewer accessible job opportunities, which perpetuates low-income status . . . Because they ride at the margins with little evidence of their plight and without a voice in the civic arena the public is oblivious to these invisible cyclists. . . “

From: Ed Beighe []
Sent: Friday, September 23, 2011 10:43 AM
Subject: Re: 1.43%

yeah, thanks for the tidbits. that makes sense, and i have noticed that myself that there are revisions.
do you take my point that news stories tend to dwell on the negatives, rather than more objective measures; so they tend to want to say that there were 10,000 bike-mv crashes WITHOUT saying or even mentioning that it would be, say, 1.5% of all crashes?
The P.R. people would call this “framing”.
FARS is fascinating; seems like 2010 should be out any day now, by the way…. or at least that seems to me to be when it comes out, late September.
when fars comes out i want to run this query: how many motorists were killed in bike crashes. Steve Magas says he has one (from 2010) in Ohio…. something along the lines of a motorists veered after bumping into a cyclist and then hit a tree or whatever and died.

To: Ed Beighe <>
Sent: Friday, September 23, 2011 8:32 AM
Subject: RE: 1.43%

I’ll see what I can find out about total State Highway System crashes only (I suspect it would show under-represented since most bike crashes are in urban areas)
btw, I calculated 1,45% and here’s a note from ADOT traffic records tech (in a message dated September 30, 2010) about using data from ADOT Crash Facts Reports –
you should always refer to the latest issue possible. In other words, for any data from 2005 to 2009, you should refer to the 2009 crash facts. For 2004 data, I would check the 2008 crash facts, and for 2003 data, I would check the 2007 crash facts, and so on. The reason is that we use the most current data available at the time of publication. So, if the number of fatalities has increased for any year since we last published the document, we will use the current data. You can note that there were 166 pedestrians killed in 2005, and this number is the most current, and any number in any previous crash facts (2005-2008) may be different (the 2006 crash facts show that 165 pedestrians were killed in 2005, and the 2005 crash facts show that 164 were killed), and I would use the number in the 2009 crash facts, and so on and so forth for other years.”
And another tidbit: 
“In terms of the crash facts data vis-à-vis the FARS data. The FARS numbers that you see are as of the date 4/15/2010. This is when FARS had their “file freeze.” In other words, in the time since we have received more information regarding the 2009 fatal crashes and the numbers have changed. What probably happened was that we received new information which indicated one of the fatalities was not legitimate, for any number of reasons. It could have been that the person died as a result of natural causes, suicide, or other things of this nature. Or maybe the person did not die within 30 days of the crash, or it is possible we discovered the crash was on private property when we received a more detailed police report, and so on and so forth. At some point the FARS file will “open” again and the numbers will be updated. I believe the 2009 data can be updated through the end of this year, and once 2011 begins, the 2009 data will be locked in place forever and ever. My guess would be at some point the FARS data will reflect the correct number of fatalities that we show. I did confirm with FARS before I ran the data earlier this month that the correct numbers were indeed 709 fatal crashes and 806 fatalities.”

From: Ed Beighe []
Sent: Thursday, September 22, 2011 1:24 PM
Subject: 1.43%

So according to my ciphering, bike-MV crashes were 1.43% of overall crashes2004-2008
i didn’t see this comparison anywhere (though i certainly may have missed it) — what fraction does the 1,089 state-highway bike-MV crashes represent; in other words are bike crashes over or under-represented on state highways?

From ADOT traffic records:
This spreadsheet shows the counts for the state highway system (SHS) and any ramps for the years 2004 through 2008. It is not completely accurate, but to get these numbers using other methods would be too time consuming. The safety data mart has a route type flag for each record, and this shows H (highway) and R (ramp) counts and excludes unknown or local roads.
IncidentYear Incident Route Type IncidentCount
2004 H 31990
2004 R 3497
2004 Total   35487
2005 H 30429
2005 R 3370
2005 Total   33799
2006 H 31382
2006 R 3804
2006 Total   35186
2007 H 33640
2007 R 4127
2007 Total   37767
2008 H 29752
2008 R 3257
2008 Total   33009
Grand 175248
So SHS:  1,089 / 175,248 = 0.62
Compared to Grand Total:  9,859 / 684,722 = 1.44
So Non-SHS might be:  8,770 / 509,474 = 1.72