Visalia, CA 93291 805 West Acequia, Suite 1A
free consultation 800 804 5476

Trad Climbing: It's More Risky Than You Think

Comments or requests for reprints may be addressed to or accessed from

Minor surgery is surgery someone else is having. – J Carl Cook
A safe climb is a climb someone else is climbing. – Robert Yoho

Who can you trust? Who is "safe"? How dangerous is climbing? How can I improve my chances of having a fatal stroke in an argument over retro-bolting at the American Alpine meeting when I am 88? Continue, gentle reader, all will be revealed.

This question of who to climb with seems different for every climber. Some people will climb promiscuously with everyone they meet, any time, anywhere. These individuals(1) don't care who you are or even what species you are as long as you can hold the rope and do the deed. Others regard climbing as so dangerous that they have confidence only in their closest friends. Sometimes this last breed is even completely monogamous, an anomaly in this odd world of odd characters and odd differences in intellect and confidence. But I say it here: first, climbing risks are very high, second, there are generally accepted standards of behavior, third, you can and should judge your partner's performance, and fourth educating yourself and eliminating problem partners reduces your risk tremendously. The standards are simple and fairly easy to learn for trad climbers (2) on reasonably safe routes with potentially good anchors. There are lessons here also for those climbing alpine or mixed routes. (3)

Climbing partners' responsibilities (4) are almost precisely analogous to those of caregivers in medical situations or of a pilot operating an airplane with passengers. All are responsible for live(s) and have duties to perform, and rules to follow, which are in the vast majority of cases simple and occasionally critical. The criteria for who should and can continue to practice anesthesia, surgery, or who can continue to work in an intensive care unit as a Registered Nurse and give intravenous medications under protocol, are well worked out and will be shown here to be closely related to judgment about who should be allowed to climb with you (of course, climbers are each responsible for the other while in medicine or flying the responsibility is one-way).

Now, how important is all of this: how risky is climbing? John Dill, in his classic article about death and injury in Yosemite (5) gives some of the best statistics available on climbing risks, see table 1. When he examined 75 incidents in the 20 years from 1970 through 1990, he found human error was probably responsible for 80 per cent of the fatalities. Good information on fatalities is easier to obtain than on accidents, which may not be reported and span a spectrum of severity. He estimates that 25,000 to 50,000 climber-days a year were involved for 2.5 fatalities on average per year.

Table 1 doesn't look so bad, but these figures are just for one day of Yosemite climbing. Taking these statistics as a crude mortality model for trad climbing (6), chances of dying when climbing actively every third weekend for two days for a year are: 35 days/year x 2.5 deaths/year / 37,500 climber-days/year or 1/429! And broken bones are 50 / 37,500 days, so chances of this for such a climber are 1/21 per year. Big, big numbers. Climb like this for 10 years and you may have 1/42 chances of dying climbing and 50:50 chance of breaking bones... To comprehend this better, see table 2 for comparison with other well-known risks. Climbing is a lot more dangerous than almost anything else you can think of. After examining these statistics, the author realized that he was subjecting himself to risk far higher than the surgical risks of his patients, and higher than surgeries he does not do because of the risk to the patient (tummy tuck).

Table 1. Yosemite Climbing Accidents (5)

Cause of Death
  • Fatalities (2.5/year)
  • Bone fractures (50/year)
  • Leader falls as per cent of accidents
  • Rockfall as per cent of accidents
  • Soloing unroped as per cent of accidents
  • Simple mistakes with gear as % accidents
In Statistical Terms
  • 1/15,000 climber-days
  • 1/750 climber-days
  • 25 per cent of fatal/near fatal
  • 10 per cent of fatal/near fatal
  • 25 per cent of fatal/near fatal
  • 40 per cent of fatal/near fatal

Table 2. General Death Rates (7)

Cause of Death

  • Accidents/unintentional injury
  • Pregnancy and childbirth
  • Heart disease
  • Cancer
  • Stroke
  • Auto accidents
  • Pneumonia and Flu
  • Homicide
  • Amateur flying small plane 100 hours in a year
  • Skydiving (parachuting) 10 jumps in 1992
  • Yearly death by traffic accident US citizens
  • Risk from tummy-tuck, (most dangerous cosmetic surgery)
  • Risk from general anesthesia, any health status
  • Appendectomy, worst calculated death rate
  • Stomach stapling for weight reduction
  • Overall general surgery risk rate (all health classes)

U.S. Fatalities / Year

  • 1/3,000
  • 1/10,000
  • 1/400
  • 1/500
  • 1/1,600
  • 1/7000
  • 1/400
  • 1/16,000
  • 1/1000 (8.6)
  • 1/6500
  • 1/7,000
  • 1/2500
  • 1/11,000
  • 1/500
  • 1/200 – 1/300 (!)
  • 1/500

And in Yosemite climbing, according to Dill 80 per cent of bad outcomes were related to human error. 80 per cent. Doing everything right would cut your accident and death rate by 4/5ths! In medicine on the other hand the vast majority of bad outcomes, including deaths are regarded as inevitable and no one's fault. One example is an accepted complication during a surgery that occurs in a statistical subset of patients. Infection, even fatal infection, occurs sometimes even when the doctor does everything right. "Negligence" however, is defined as when a doctor doesn't perform according to accepted standards (8): the human error that Dill refers to above. The medical board in each state has the responsibility to evaluate complaints and determine if the doctor has the wherewithal to continue to practice medicine, or if his license should be discontinued or monitored. The discipline process in regard to a physician is complex and the issues are sometimes obscure, but let it suffice to say that a lot of people are watching us and we have less responsibility statistically than a climbing partner and far less statistical likelihood of negligence.

And registered nurses routinely get fired for medication errors. Just one medication error, sometimes. They have been trained for years to be careful and double-check meds. Small plane pilots have roughly the equivalent risks per hour (8.5) as Yosemite trad climbers. These people have been trained and tested exhaustively, their airplanes are subject to extensive maintenance requirements, and what they are doing is acknowledged (by everyone but the pilots themselves!) as very hazardous.

As the first author put together the information for this article, he found that his own safety and rescue skills were rudimentary, despite climbing experience of over 35 years (see box). He couldn't do a mechanical advantage hauling system, he never carried a knife, and there were embarrassing lapses in some of his rappelling techniques. After some reading and changes in some habits, he has tried to be a safer climber and a better judge of his partners.

Abbreviated Climbing CV: Robert Yoho, MD

First ascents:
  • El Matador, Devils Tower, 11a
  • Carol's Crack Devil's Tower 10d
  • Various Joshua Tree short free climbs
approx. year
  • mid 70's
  • mid 70's
  • mid-late 70's
Selected free climbs:
  • Naked Edge El Dorado Cyn, Boulder
  • Astroman, Crucifix, W. Face El Capitan, Rostrum
  • Chouinard Herbert (Yosemite)
  • Beggar's buttress free except last pitch (Yosemite)
  • late 70's
  • mid 80's
  • 2004
  • Nose, Leaning tower
  • Half dome regular route
  • Half dome in day
  • late 70's
  • mid 70's
  • 2004

Each climber has to be his own "medical board" in regard to himself and his partners. There is no formal licensure, but your rough guidelines are right here in this article. If someone with years of experience and who is assumed to be competent violates some basic principle, you should terminate his license to trad climb with you. They either are 1) too dumb to understand the systems' importance and intricacies or 2) messed up emotionally by their personal circumstances so their judgment is haywire or 3) unconsciously suicidal, homicidal or depressed or 4) freaked-out by the climb so they can't perform or 5) on drugs 6) doesn't get along with you and so is acting out or... a host of other possibilities. And you don't want to climb trad with an un-athletic klutz who is smart but may make a coordination mistake that results in a big problem. Can your partner habituate to the risks and apparent risks of climbing, and function with a clear head through these pressures? Here's an example of someone who is no dummy, and a famous mountaineer, but has one type of problem that should rule him out as a partner. He says, "both early and late in my mountaineering career, my greatest failing has stemmed from an apparently innate mechanical incompetence. Gear of all kinds baffles and infuriates me, and I am incapable of repairing the simplest device of any sort, let alone contemplating how it works." He then goes on to describe fatalities that were at least partly related to gear screw-ups. This guy was educated to the best climbing standards of the day (1960's) and a very competent climber physically who put in world-class routes but from what he writes, it isn't completely clear he was always tying bowlines at anchors. And he still (in 2005) seems to be in the dark about the fact that two half hitches transpose into a square knot, not a granny. (8.6)

And you'd better have some idea of your own physical and emotional state when climbing. If you aren't sure you know what you are doing, you'd better do some reading, take a course, and practice in a controlled environment. It's bad for your health to be ignorant or distracted. You can't evaluate your behavior or your partner's.

What is the cause of a partner's error(s)? It doesn't really matter. This person is putting your life in big, big danger. Friendship is not even a consideration. To paraphrase Mario Puzo in The Godfather, any mistake of a serious kind is both personal and business. The important question is: do you give them another chance? As with medical personnel, the answer is: maybe. It depends partly on how they respond to your explanation of the problem. If they readily understand and admit the error and there's some sort of good excuse (two days of no food or water for example), maybe.

What is a mistake? Protection, belay, anchor, and other climbing standards have been explained in elegant and interesting ways elsewhere (8.75) and it is beyond the scope of this article to give more than a few obvious examples. Additionally, judgment differs depending on the type of climbing you are doing and your personal definition of the risks you are willing to take for your sport. Know the risks of what you do, and understand what you personally define as insanity. The first author, for example, feels that risks of alpine or altitude climbing are unreasonable. Himalayan climbing, of course is very risky; "a rigorous German study ... concluded that on any given expedition to an 8000-meter peak, a climber stands a 1 in 34 chance of dying" (8.6)). Much of the following should be universal, although there is quibbling about some of it. (8.65)

Rigging the anchor so if one point fails, the whole thing unthreads is grounds for licensure revocation. No second chances. Using the Euro death knot (9) probably should be a no-no. Running it out off the anchor after the belayer explains that fall factor reducing pieces might put your partner in "sport climbing only" category (10). If your friend runs it out on lead after pieces have been placed and the anchor is secure: you might consider it his problem if he breaks his head. Refusing to use a Cinch or Grigri for belay (11) probably increases your risks. Should you climb with Dean Potter? Nope. We admire the son of a bitch, but he doesn't seem to have it in his head that he needs a minimum of two anchors at all times to keep him alive (12). And your partner should not rappel without a backup friction knot.

Work on teaching your partners mechanical advantage hauling systems. You will be safer for it.

The take-home message: you have to educate yourself constantly and keep your eyes open. And you can't let "friendship" stand in the way of staying alive. Trad climbing is a MUCH more dangerous game than most people realize, with generally even more responsibility than the practice of medicine. You wouldn't let a hack operate on you, don't allow anyone second rate to be your trad partner.

Stop climbing with anyone, no matter how much you like them, after two significant judgment errors, especially in one day. Go sport climbing with them, but not trad. You don't want a spinal injury, and you don't want to bother with a rescue. Just pray to your gods for a few more years of good health.

End notes

(1) Credit goes to Bill Wright and Hans Florine in their wonderful book Speed Climbing for this colorful terminology.

(2) Although sport climbing bolt-to-bolt is risky compared to a meal at KFC®, the risk is generally small enough that it is hard to quantify. In author's opinion, in this arena, you (alone) can generally control the entire human error risk to yourself by judicious—read alert and paranoid—observation, plus universal use of a Grigri® or Cinch®.

(3) Which are orders of magnitude more complex and hazardous and require many other skills and judgment qualities. Their screening and evaluation process for partners must be more exacting. Because of the risks accepted--witness the graveyard in Chamonix--the homoerotic quality the Mark Twight has with his partners is not at all misplaced, even if Twight is indeed heterosexual. (The author makes no speculation and has no knowledge regarding this issue, and considers what Twight does in his time outside of climbing to be entirely his own affair). In this arena, you actually have to get to know your partner. It's a requirement.

(4) Of course, all bets are off for a beginner or guide-beginner relationship. Beginners must be supervised with an eye like a hawk, and well-trained, experienced guides should have a paranoia level higher than a rat in the snake cage. If the beginner is injured or worse, the guide is ethically responsible. This is true even if the climber guided is experienced. The herein is about climbing (or not climbing) with someone who is your peer.

(5) Dill, John "Staying Alive" accessed August 17, 2005, at

(6) US trad climbers in other locations probably have lower average risks per climbing day, so these statistics may overstate USA trad climbing morbidity and mortality somewhat.

(7) Yoho RA, Romaine JJ, O'Neil D. Review of the liposuction, abdominoplasty, and face-lift mortality and morbidity risk literature. Dermatol Surg. 2005 Jul;31(7 Pt 1):733-43; discussion 743. For many more statistics of this kind and explanations of risk in more detail, you may download this article from

(8) When his patient is injured as a result, there may be grounds for a lawsuit. The full definition of negligence: "the omission to do something that a reasonable person, guided by those considerations that ordinarily regulate the conduct of human affairs, would do, or the doing of something that a prudent and reasonable person would not do. Negligence is not an absolute term, but a relative one, and whether or not a particular act or omission constitutes negligence depends by definition on the particular circumstances, considering especially the time, place, and persons involved. A determination of negligence is a legal conclusion that can only be arrived at by a court of law." Gross negligence and incompetence are even worse. Gross negligence has been defined as "an extreme departure from the ordinary standard of conduct, as an entire failure to exercise care, as the exercise of so slight a degree of care as to justify the belief that there was an indifference to the interest and welfare of others, and as that want of care that raises a presumption of conscious indifference to consequences. A determination of gross negligence is a legal conclusion that can only be arrived at by a court of law." (source of these definitions:, accessed August 17, 2005.

(8.5) Fallows, J. Free Flight. New York, NY: Public Affairs New York; 2001:59.

(8.6) Roberts, David, On the Ridge Between Life and Death, Simon and Schuster, NY, NY 2005, p134-5 and 145.

(8.65) While reviewing the available information for this article, one author (RY) was surprised at his own lack of sophistication, and commenced an extensive review of current standards.

(8.7) Fasulo, David. Self Rescue. Falcon Press, 1966. Robbins, Royal. Basic Rockcraft. Long, John. Climbing Anchors. Etc.

(9) see (accessed August 17, 2005) for details of why overhand loop without a backup may be an unwise choice for joining rappel ropes.

(10) (further details of this with referenced source would be helpful, I just heard a rumor) Climbers on Middle Cathedral last year had three camming units deform and pull out of granite and they got the quick trip to the Valley floor after an apparent factor 2 fall. We do not know whether the pieces were equalized, but they were all three apparently in there pretty good because the units were ruined. My paranoia about the leader falling on the anchor is unlimited, even with a good anchor.

(11) It is my firmly held conviction that a Cinch or Grigri with the brake hand off the rope is safer for the climbing team as a whole than the most experienced and reliable belayer in the world using a plate or other passive device. Why? If a rock hits my compulsive plate belayer and kills her, we find that I fall and die also. I live to self-rescue if she's using the Cinch and I of course am able to additionally rescue her if she is merely unconscious, all because the device works when the belayer is inattentive. All this and most of us have seen people dropped 20 feet by accident because some experienced hot shot took his hand off the belay rope using some stupid tuber or plate. My opinion: none of these plates are as good as a Munter knot for belaying except perhaps a Reverso.

(12) For the beginner, and for some advanced climbers I know: the most basic precept in climbing is that you need two absolutely bombproof and separately constituted items to attach you to the rock, and if you don't you'd better hold on tight because you never trust just one except the Rope and the Rope is sort of a God you believe in, because rope failure is almost unheard-of.

Do You Have Any Idea At All What You Are Doing...

A Continuing Education test for Trad Climbers. (MAY NOT APPLY TO ALPINE, where other considerations such as speed, weight, and equipment limitations are more significant to overall safety.)

1) What is the most basic principle in protection in rock climbing roped climbing? Hint: our hero Dean Potter violates this principle routinely but is still alive because House odds advantage hasn't caught up with him yet.

2) Explain what fall factors are and how this relates to leading off anchors and how the anchors are loaded if a fall occurs.

3) How many equalized #2 Camelot equivalent do you need for adequate anchor in Yosemite granite?

4) The following are rappel anchors. Explain what, if any, the problems are with the systems the way they are set up. (diagrams show various systems with one link in rappel chain at some point.)

5) How could you correct problem(s) with any of the above rappel setups. (leave a biner or thread a runner to make double systems at the only point which is a single system)

6) Explain the differences between a single rope system, double rope, and twin rope systems.

7) Diagram 3:1, 5:1 and 7:1 mechanical advantage hauling systems. (No, this is not for extra credit. You are a trad climber and need to be able to haul your unconscious partner up to a ledge).

8) What is a euro death knot and what are it's advantages and disadvantages? For extra credit, in what country is it taught as the brainless standard? Hint, it's not in Europe. (New Zealand)

9) Examine the following two anchors. Describe the problem(s) if any with each. (2 bolts, one double sling, one example with sling clipped with self-equalizer method, but with one bolt going out, other would be shock loaded. This has single locker clipped into rope that goes to climber, a barely acceptable option. Other system, unacceptable, has Reid Malenbaum method of clipping both slings in middle, one blows and the whole thing unthreads.)

10) Name four methods for backing up rappels. (knots in end of rope, autoblock, Kleinheist or prussic knot with extender, sliding ascender down the rope above the rappel system, partner at bottom of rappel holding on to rope to tighten it if out of control event happens. Simul rap on separate ends with 6 foot leash tying partners together.

11) Show how to tie autoblock and Kleinheist.

12) You drop your jumars but have to go down a fixed line to free it. How do you get back up? (three answers) (alternate prussics with backup knots tied, Bachman with backup knot tied, prussic or Bachman with foot in loop of rope or ______ knot for your foot. Any answer that does not include backup knot, no credit and two points off).

13) Do you have a knife on your harness or in your pocket when you trad climb?

14) Do you wear a helmet?

15) See the following diagrams. Is 1) more reliable or 2) and explain why. (Tied off piton with girth hitch versus carabiner clipped through eye of piton, which protrudes from crack half it's length.)

16) What can be relied on as single system to support your life in climbing? (Single rope UIAA rated, carefully observed to be not running over an edge. Belay device, locking carabiner and harness loop when you are belaying partner. Many guides rely on locking carabiners as single system for clipping into an anchor, but authors believe these should be backed up.)

17) By what percent do you improve your chances of living through the climb if you and your partner both use self locking devices for belaying the leader, i.e. Cinch or Grigri? (authors estimate 5-10 per cent, even with best belayer who ever existed: see rationale in footnotes).

18) What is rope creep on rappel and what are potential consequences? (Thicker rope tends to stretch less and the thinner rope goes through the anchor, thus when you get to the end of the lines there is a potential for rapping off the shorter (thicker) line if you haven't tied knot in ends of rope.

19) Describe munter-mule knot and what it is used for. (diagram. Used with prussic, kleinheist, or other friction knot to take tension off a weighted rope for various purposes, such as escaping the belay to rescue or go for help.)

20) Name the advantages and disadvantages of belaying from the power point, and describe exactly what this is. (Advantages: easy escape from belay. Be careful with passive device: you need a redirect of belay rope going from the device to belay hand for proper safety.)

1-4 right, you are qualified to toprope at the local gym.
4-8 you have lead privileges at your gym. Go outside only to sunbathe.
9-12 right: you may take a climbing class from an ACGME approved guide.
12-18 right: toproping outside OK, don't lead.
18-24 You may lead, but be careful out there.