Tolerance is how far a plate’s real weight is allowed to differ from the number stamped on it, written as a ± percentage or a ± number of grams. Economy cast iron often runs a few percent off, and a published tolerance of about two percent is good for a budget plate. At the opposite end, calibrated competition plates are held to roughly ±10 grams.The gap between those two figures is enormous — and understanding it tells you how much to care about the number on your own plates.
What the number actually means
When a maker prints a tolerance of, say, ±2%, they’re telling you the plate can weigh up to two percent more or less than its label and still count as in-spec. On a 45-pound plate that’s a window of roughly 44.1 to 45.9 pounds. A tolerance in grams works the same way but as a fixed amount rather than a percentage — ±10 grams is ±10 grams whether the plate is 5kg or 25kg. Neither means the plate is “wrong.” It means real manufacturing has limits, and the tolerance is the honest statement of them.
The usual ranges
As-cast economy ironis the loosest, simply because pouring metal into a mold can’t control mass tightly — a few percent is normal, and around two percent is a good result for the price. Machined plates tighten that up, because CNC facing and boring remove metal to a target; the connection between finishing and accuracy is the whole point of our machined versus cast guide. At the top sit calibrated competition plates, held to about ±10 grams so that a record set on one is legitimate anywhere. That extreme precision is compared against rubber in our calibrated versus bumper plates guide. Most home lifters live comfortably in the cast-to-machined range.
Why it matters
A single plate a little heavy or light is meaningless. The reason tolerance is worth a thought is that errors accumulate. Load five plates per side, each two percent heavy in the same direction, and the bar can be a few pounds off what the math says. That won’t hurt your training, but it matters in two situations. First, balance: if a plate and its pair differ, the bar is heavier on one side, and matched pairs keep it even. Second, PR accuracy: if you’re chasing a true number to the pound, you want to know the iron is close to honest. This is also why fractional platesexist — when small, accurate increments matter, you want them made to a tight tolerance too.
How to sanity-check plates at home
You don’t need lab gear. Put a plate on a decent bathroom or postal scale, compare it to the stamped weight, then weigh its pairthe same way. You’re checking two things: is each plate roughly right, and do the two that should match actually match? If a matched pair differs noticeably, the simple fix is to load them on opposite ends so the imbalances cancel and the bar stays even. Don’t obsess over a few grams — home scales aren’t precise either — but a plate that’s wildly off its label is worth flagging, and a machined line like the standard plates we reviewed will usually check out closer than bargain-bin iron.
How much should you care?
Be honest about your goals. For general strength and hypertrophy training, a couple of percent is invisible— your muscles can’t tell 225 from 227, and progressive overload works fine either way. Tighten your standards if you compete, if you like your logbook to mean something to the pound, or if you simply value precision. When you’re shopping, tolerance sits alongside diameter, hole size and finish in the overall plate-buying decision— a spec to read, not a reason to overspend. Knowing what the number means is enough to stop it from misleading you.
Where tolerance shows up in practice
A few concrete cases make the idea stick. Say you load a bar with four 45-pound plates per side, each running one percent heavy in the same direction — that’s roughly a pound and a half of hidden extra weight, small but real if you’re grinding a top single. Now imagine one of those plates is a percent heavy and its pair is a percent light: the bar is subtly heavier on one end, which you might feel as a tilt on an overhead press or a bench lockout. Neither situation is dangerous, but both are the kind of thing tighter tolerance quietly prevents. It’s also why the habit of buying plates in matched pairs from one linepays off — plates made together tend to share the same small bias, so they cancel rather than compound. The same logic scales down to the smallest jumps: when you’re adding just a pound or two with fractional plates, a loose tolerance would swallow the whole increment, so good fractionals are held tight. And it scales up to the bar itself, which has its own weight and rated capacitythat you fold into the total. Tolerance isn’t something to fear — it’s just a number to read and account for.
Sources
- Manufacturer-published weight-tolerance conventions for economy cast iron (a few percent, with roughly ±2% considered good) and calibrated competition plates (about ±10 grams), as reflected in plate spec listings linked from our reviews.
- General plate-accuracy and balance principles as covered across our plate guides and our methodology.