American Alpine Jounrna and Accidents in North American Climbing

Equipment Failure — Ascenders Came Off Rope, Fall on Rock, Inadequate Self-Belay, California, Yosemite Valley, El Capitan

  • Accident Reports
  • Accident Year:
  • Publication Year: 1998


California, Yosemite Valley, El Capitan

On May 27, Robert Jatkowski (30) was cleaning pitch 28 of the Shield, while his partner, Uwe Reissland (38) waited at the belay above.

After finishing a long left-facing vertical corner, Jatkowski reached a Lost Arrow pi- ton about one meter below a roof. From the Lost Arrow the rope ran up and left at 45 degrees to a fixed piton under the roof, a meter left of the corner, then traversed several pieces further left.

Preparing to pass the Lost Arrow and clean it, Jatkowski detached his upper ascender with his right hand, reached up and left with it as far as he could, and clipped it to the rope below the fixed piton. Then he pulled down on the rope below the lower ascender, also with his right hand, and released its cam with his left. His intention was to un-weight that ascender and let rope slide through it until his weight came onto the upper ascender; this would swing him to the left under the fixed piton—a typical move when traversing.

The next thing he knew, both ascenders were off the rope and he was falling. He was not tied in short and was near the top of the pitch, so a long loop of rope hung below him. Most of the fall was free, but his left foot struck a ramp after 10-15 meters. He estimates the total fall-length was 35-40 meters.

After coming to a stop he checked himself, decided he had no serious injuries, rerigged his ascenders on the rope, and climbed up again. His left foot hurt but he was able to finish cleaning the pitch. He was also able to follow the next pitch, the last on the climb, using only his uninjured leg, but the 2nd class summit slabs brought him to a halt.

Jatkowski waited at the top of the climb while Reissland hiked to the Valley and notified the NPS. Rangers flew to the summit, splinted Jatkowski's leg and helped him hobble 400 meters up the hill to the LZ. At the clinic he was diagnosed with a fractured left ankle.


Jatkowski’s upper ascender, operated by his right hand, was a right-handed Petzl. The lower one was a left-handed model almost identical to the Petzl, but made in Czechoslovakia. Each ascender was rigged in typical big-wall fashion, with a daisy chain to the harness and an étrier to a foot.

Neither climber saw how the ascenders came off the rope, but it seems unlikely that a correctly attached Petzl—with the safety latch fully engaged—could escape by itself. So Jatkowski probably did not completely reattach the upper ascender—more likely since reaching up and left with his right hand was an awkward move. The spring torque in the cam will usually swing the body of an ascender parallel to the rope by itself, allowing full engagement of the cam and the safety, but tension from the daisy and the etrier can easily prevent this; sometimes just the weight of the slings are enough. In these cases the climber must manually rotate the ascender into position.

Since the ascender was right-handed, the cam, safety, and rope channel now faced toward the rock, away from Jatkowski. He remembers that he did not check the ascender after attaching it, so an incomplete attachment would have been easy to miss. Jatkowski opened the cam on the lower ascender by pushing back on the thumb bar attached to the safety latch. This moves the cam away from the rope, allowing the rope to slide in the channel, but it also begins swinging the safety out of the locked position. If the rope is pressing against the cam at this time it can push the cam all the way open and escape from the device. This is possible with both the Petzl and the Czech model, although the latter seems more prone to it because of the shape of the safety latch.

Jatkowski, Reissland and I rigged a full scale model of the problem on the wall of a building, and I simulated Jatkowski's action while wearing his harness and ascenders as he had rigged them. When I re-attached the upper ascender incompletely to the rope it did not automatically rotate into the locked position, being held back by the daisy and étrier.

Its tempting to assume that we know the answers, but our evidence is really only circumstantial. We do know, however, that Jatkowski would not have fallen so far had he tied in short to his rope (just below the lower ascender) before the traverse. He is very experienced, including on Yosemite walls, and has tied in short in the past, but it just did not occur to him in this case. With the summit only a pitch away, he felt he may have been climbing hastily and not concentrating on the task at hand.

Several injuries and deaths have happened this way in the park, with several models of ascenders (and almost always on traverses). In every case, tying in short would have prevented the accident, and checking the ascender after attaching it would probably have prevented the incident in the first place. (See, for example, ANAM 1987, Schrattner, El Capitan.) (Source: John Dill, NPS Ranger, Yosemite National Park)

This ANAM article has been reformatted into HTML. Please contact us if you spot an error.