Climbing Knot Guide

The standard tie-in knot. Almost every climber in the world uses this to attach the rope to their harness.

1. Tie a loose figure-8 about a metre from the end of the rope.
2. Thread the tail through both harness tie-in points, the same path the harness loop takes.
3. Follow the original 8 back through, keeping the strands parallel and untwisted the whole way.
4. Dress it flat and tighten all four strands individually.
5. Leave a tail at least ten centimetres long.

How it fails: It is easy to tie a figure-8 that looks right and is missing a pass. Trace the whole knot with a finger before you leave the ground, and check your partner as they check you — the partner check exists because this knot fails silently.

A fixed loop anywhere in the rope, for clipping into an anchor or building a master point.

1. Fold a bight of rope where you want the loop.
2. Tie a figure-8 in the doubled rope.
3. Dress it so the strands lie flat and parallel.

How it fails: It welds itself shut after heavy loading and can be genuinely difficult to untie. Where you will need to release it under or after load, an alpine butterfly is the better choice.

An alternative tie-in that unties easily after being weighted, which is why some sport climbers prefer it after repeated falls.

1. Make a small loop in the standing rope.
2. Pass the tail up through the loop, around behind the standing part, and back down through.
3. Dress and tighten, then back it up.

How it fails: A plain bowline can shake loose when unloaded and cycled, which is exactly what happens on a long walk to the anchor. It must be backed up — a double bowline with a stopper, or a Yosemite finish. Untied bowlines have killed people; the knot is fine and the missing backup is not.

Attaching yourself to an anchor with a length you can adjust one-handed without unclipping.

1. Make two loops in the rope, both crossing the same way.
2. Slide the second loop behind the first.
3. Clip both loops into the carabiner and pull both strands tight.

How it fails: It must be loaded across the spine of the carabiner, and the gate must be locked. A clove hitch can slip a little under a shock load before it grips, so it is an anchor attachment rather than a critical load-bearing knot.

Belaying or rappelling with nothing but a locking carabiner. The knot that gets you out of trouble when a belay device goes down the hill.

1. Make a loop in the rope and fold it back on itself so the strands cross.
2. Clip both strands into a large pear-shaped locking carabiner.
3. The hitch flips through the carabiner as you switch between paying out and taking in.

How it fails: It twists the rope badly, so expect kinks. Use a pear-shaped locker with plenty of room, and never a small D — the hitch needs space to flip, and a cramped carabiner is where it binds.

A friction hitch that grips the rope under load and slides freely when you release it. The core of every self-rescue and hauling system.

1. Tie a loop of cord with a double fisherman's bend.
2. Girth-hitch it around the rope, then pass it through itself two or three more times.
3. Dress every wrap so they lie neatly side by side without crossing.

How it fails: The cord must be meaningfully thinner than the rope — around 5 to 7 mm on a standard single. Too thick and it will not grip. It also grips poorly on wet or icy rope, which is exactly when you need it, so add a wrap in those conditions and test it before you trust it.

A friction hitch that grips in one direction only and releases more easily than a prusik. Works with webbing as well as cord.

1. Wrap the cord loop around the rope three or four times, all in the same direction.
2. Pass the long end through the short end.
3. Load the long end downward.

How it fails: It is directional and will not hold if loaded the wrong way. Check which way it grips before you commit, because a klemheist tied upside down looks exactly like one tied correctly.

A rappel backup below the device. It grips when you let go and releases easily under a hand.

1. Wrap a short cord loop around both rappel strands four to six times.
2. Clip both ends to a leg loop.
3. Keep a hand on it, sliding it down as you descend.

How it fails: It must sit below the rappel device, never above. Above the device it can jam out of reach and leave you hanging with no way to release it — a situation that requires a full self-rescue to escape.

Joining two ropes or cords permanently. The standard way to close a prusik loop.

1. Lay the two ends alongside each other, pointing opposite ways.
2. Wrap one end twice around the other and back through both wraps.
3. Repeat with the second end in the mirror image.
4. Pull the two knots together so they seat against each other.

How it fails: It jams solid after loading and is often impossible to untie, which is a feature for a prusik loop and a problem anywhere you want the rope back. It is also bulky and snags when pulled over an edge, so it is not the first choice for joining rappel ropes.

Joining two rappel ropes. The flat profile slides over edges and rolls over rock far better than a bulkier knot, which is the whole reason it is used.

1. Lay the two rope ends side by side.
2. Tie a simple overhand knot in both together.
3. Dress it perfectly flat with no crossed strands.
4. Tighten each of the four strands separately, hard.
5. Leave tails of at least 30 centimetres.

How it fails: This knot can roll along the rope under load, and each roll consumes tail. That is what the 30 cm tails are for. Never tie the flat version of a figure-8 instead — it looks tidier, rolls far more readily, and has killed people. If in doubt, add a second overhand behind the first.

A loop in the middle of the rope that can be loaded from any direction and still unties after heavy loading. The knot for isolating a damaged section or tying into the middle of a glacier rope.

1. Wrap the rope twice around your hand.
2. Pass the outer wrap over the other two and back under.
3. Pull the resulting loop through and dress it.

How it fails: It relies on being dressed correctly and a badly dressed butterfly is noticeably weaker. Learn to recognise the finished shape — three strands entering, three leaving, symmetric — rather than just following the motions.

Attaching a sling to a harness or around an object, quickly and with no hardware.

1. Pass the sling around the object.
2. Thread one end through the other.
3. Pull tight so it seats evenly.

How it fails: A girth hitch weakens a sling substantially — commonly quoted at around a third — because of the sharp bend it forces. That is acceptable for a personal tether and not acceptable where the sling is load-bearing on an anchor.