You test one-handed and the number lands well under half of your two-hand lift. You check the strap. You check the edge. You wonder whether you've been kidding yourself the whole time.
You're not broken. The arithmetic that says one hand should be half of two hands assumes something about the two-hand lift that isn't true.
What a two-hand lift is actually measuring
When both hands are on the block, nothing enforces an even split. The stronger side takes what it can, the weaker side contributes what it must, and the total that leaves the floor is a single number that hides the arrangement entirely. That is the point our testing post makes when it explains the conversion from an old two-arm hang: you can divide the total by two to estimate per-hand load, but treat it as an estimate rather than a test result, because leaning on the stronger side means a converted number usually flatters you slightly.
Add to that everything a two-hand setup does for your body. Two contact points make a stable, symmetrical position. Your torso doesn't have to resist rotation, your shoulders share the job, and the stance is more stable over the load. Take one hand away and the whole system gets less efficient before your fingers are even part of the conversation.
The bilateral effect, at the depth we can support
There is also a documented tendency for people to produce less total force when both limbs work together than the sum of what each produces alone. It has been observed across sports and body parts, and asymmetry between limbs is documented in climbers specifically, which is part of what the citation in our post on the unlevel edge is describing.
We're not going to give you a percentage for how big that effect is in climbers' fingers, because we don't have a source we trust for the size of it. Direction, yes. Magnitude, no. If someone quotes you a tidy figure, ask where it came from.
The research has the same problem. A systematic review of finger-strength testing found the tests themselves highly repeatable, with reliability coefficients between 0.85 and 0.99 for maximum strength on a fixed edge, and then noted that several of the studies testing both hands did so one hand at a time rather than simultaneously, which makes the result less representative of how force actually gets produced in climbing. Even the people whose job is standardising these tests treat a two-hand arrangement and a one-hand arrangement as measuring something different.
Two ledgers, and what each one is good for
The two-hand lift is a loading tool. It lets you handle heavier absolute loads with a stable body position, it is easier to set up and repeat, and it is comfortable enough to run for volume. If the session's job is to put load through your fingers, this is the efficient way to do it.
Half of your two-hand lift is not your one-hand lift.
The one-hand lift is a measuring tool. It's the version that answers questions a combined number can't. Which hand is stronger, and by how much, which our companion post on left-right asymmetry takes up in detail. What your actual per-hand capacity is, without the arrangement hidden inside a total. And it's closer to climbing, where your hands are rarely doing the same job on the same hold at the same instant.
Our testing protocol is one-armed for exactly that reason: the two are rarely equal, and averaging them hides the gap.
Warm up first and stop on pain. Full safety guidance.
Rules for keeping both
- Never convert one into the other and log it as a result. Halving a two-hand number is a rough estimate, and it is a generous one. If you want a one-hand number, test one hand.
- Keep separate columns. Two-hand loads and one-hand loads are different lines in the log, the same way a half crimp and a drag are different lines. Mixing them turns a change of method into fake progress.
- Keep the setup fixed within each ledger. Same edge, same grip position, same stance, same way of standing over the load. All the hygiene from the testing post applies to both. This is where those reliability figures come from: they belong to protocols that hold edge depth and grip constant, not to the test in the abstract.
- Expect the one-hand number to be the honest one. It is smaller, it is more annoying to produce, and it is harder to flatter yourself with.
How this shows up in a week
How much of each you run is a programming question, and it is one we're deliberately not prescribing here, because a specific split of sets between one-hand and two-hand work is coaching judgement and we would be inventing it.
What the two tools can claim is what they see. If you never test one hand at a time, you have no idea whether you have an asymmetry, and you won't find out until a move that loads your weaker side stops you. If you only ever lift one-handed, you're giving up the most efficient way to put real load through your fingers. Most people who own both end up using two hands to train and one hand to check, which is less a recommendation than an observation about what each is good at.
The tool
Both jobs run off the same kind of edge, and the only hard requirement is that it doesn't change between tests. Pick one edge as your reference, the Dual Edge Block or whichever you already train on, and keep every number that matters on that edge.
The bottom line
Your one-hand lift isn't a failed version of your two-hand lift. A two-hand number hides how the load was shared, gives your body a more stable position to produce it from, and flatters you a little when you halve it. A one-hand number is smaller and more honest, and it's the one that can find an asymmetry. Keep both, keep them in separate columns, stop doing arithmetic between them, and let each one answer the question it is actually able to answer.