There is no honest climber push-pull ratio. Gym culture has them: bench to row, press to pull-up, some tidy figure someone read once. Those come out of general strength training and they were never built for a sport where one side of the balance is trained thousands of times a week at bodyweight and the other side barely exists.
How unbalanced am I? Give me a number.
The ones that circulate in climbing media are mostly invented. Quoting a bench-to-row figure at a climber borrows authority from a context that doesn't apply. What replaces a ratio is less satisfying and more useful: measure your own pushing capacity properly, then watch it over time against your own pulling. You're not looking for a target. You're looking for a direction.
What to measure
Pick two or three tests, keep them fixed forever, and choose from movements you already use for antagonist work rather than inventing a battery:
- A max quality set of parallette push-ups. Full range, chest to hand height, no sagging through the middle, controlled tempo. The set ends when the quality goes, not when you fail.
- A dip or a dip hold. Either a max quality set of dips, or a timed hold at the bottom position if your shoulders prefer that. Pick one and stay on it.
- An L-sit or tuck hold for time. This is the straight-arm scapular quality that carries over to steep climbing, and it is the one people are worst at and quickest to cheat.
And a pulling number you almost certainly already have: your max lift, your pull-up max, whatever you already track. You're not converting between the two. You're watching whether one of them moves for a year while the other stays flat.
Testing rules, borrowed wholesale
The hygiene from our finger testing post transfers to pushing without modification, and pushing tests are easier to fool yourself on because there's no plate to be honest for you.
- Same setup, every time. Same parallettes, same hand spacing, same tempo, same depth.
- Define quality before you start. Write down what a rep looks like. Any rep that doesn't look like it doesn't count, and the set ends there.
- Warm, and not at the end of a session. A cold pushing test reads low and a fatigued one reads lower.
- Log the conditions along with the number. Date, tempo, spacing, how you felt, all of it, exactly as our log post argues for finger training.
Reading the result
Since there are no bands to place yourself in, read it as a trend, which is how our log post says to read any noisy measurement anyway. Retest at the start and end of a training block, not weekly. Two useful shapes:
Pulling numbers up, pushing numbers flat for a year. That's the imbalance arriving in your own data instead of as a general claim. It's the case for giving pushing a standing slot.
Both moving, slowly. Nothing to fix. Your modest amount of antagonist work is doing the job it was supposed to do, which was never to make your push impressive.
What you can't conclude from any of this is a risk number. The hard experimental evidence that antagonist training prevents climbing injuries is thinner than the confident advice you'll read online. A tracked pushing number tells you whether you're training the side you say you are training. It doesn't tell you your shoulder is safe.
Acting on a lopsided answer
The dose is small and already written down in the push-day post: a couple of short sessions a week, quality sets rather than a grind, progressing by harder shapes rather than endless reps, and treating the whole thing as joint insurance rather than performance training. Adding your first two pushing sessions a week is the entire intervention. If you find yourself building a chest day, you've overshot the point.
The equipment note
Parallettes are the tool for most of these tests because they let you press with the wrist in a neutral position, which matters if your wrists are already busy being crimped on. Ours are 2 Inch Parallettes: wooden, with non-slip rubber feet, built for push, core, and shoulder stability across planche work, L-sits, and antagonist training.
The two inches are the point rather than a rounding of the number. Most parallettes are made at 1 5/8 inch, and the thicker bar is more comfortable in the hand under load, which matters most in exactly the positions you're about to hold for time. If you're testing L-sits and holds regularly, the handle you're gripping isn't a detail.
The bottom line
There is no honest push-pull ratio for climbers, so stop looking for one. Pick two or three pushing tests, define what a quality rep looks like, run them at the start and end of a block under identical conditions, and compare the trend against the pulling numbers you already track. If pushing has been flat for a year while everything else climbed, you've found the imbalance without needing anyone's ratio. Then do the small amount of work the push-day post already prescribes and get back to climbing.