The standards that matter are ISO 6789 and the traceable work NIST performs on torque realization. Calibrate yearly, stay out of the bottom 20% of your wrench’s range, and use correct technique every time you pull the handle.
TL;DR:
- A ±4% rating permits a 4 ft·lb error at 100 ft·lb; compare that margin with joint tolerance, and check the separate rating for counterclockwise use.
- Choose an ISO 6789 Part 2 calibration for traceable results; Part 1 confirms tolerance but reports no measurement uncertainty.
- Keep working torque above 20% of the wrench’s rated range; for lower values, request calibration at that point or use a lower range wrench.
- Recalibrate manual wrenches every 12 months, sooner after heavy daily use, a drop, storage under compression, or an unexplained field check bias.
- Use the marked grip, pull smoothly, stop at the click, and account for extensions because hand position and accessories can change delivered torque.
Table of Contents
- What an Accuracy Rating Actually Tells You
- Why the Bottom 20% of Your Wrench’s Range Is a Problem
- Calibration Standards and How Often to Recalibrate
- How Your Technique Changes the Torque You Actually Deliver
- Field-Testing Your Torque Wrench Between Lab Visits
- When ±4% Accuracy Is Good Enough for the Job
- How We Build Torque Accuracy Into Our Tools
- What Actually Matters Most for Accuracy
- Pick the Right Torque Tools for Your Shop
- FAQ
- Sources
- Key Standards, Labs, and Testing Articles
What an Accuracy Rating Actually Tells You
At 100 ft·lb, the error widens to a 4 ft·lb spread in either direction.
Accuracy and precision aren’t the same thing. Accuracy tells you how close the tool gets to the true value; precision (often called repeatability) tells you how consistent it is click after click. A wrench can be precise and still be inaccurate if it’s drifted off calibration but keeps clicking at the same wrong number every time.
Direction matters too. Many click wrenches carry different ratings for clockwise and counterclockwise use, and testing on entry-level tools found models holding ±4% clockwise but a looser ±6% counterclockwise.
- A ±4% rating at 100 ft·lb means a 4 ft·lb window above or below your target.
- Clockwise and counterclockwise accuracy are often rated separately on the same tool.
- Unit-to-unit variation exists even within the same model line, especially on budget tools.
Many entry-level wrenches meet their advertised ±4% clockwise accuracy when new, according to independent user testing, though they tend to show more drift and inconsistency over time than professional-grade tools.
Why the Bottom 20% of Your Wrench’s Range Is a Problem
Torque wrenches get tested at three points: 20%, 60%, and 100% of their rated range. Those three readings are what manufacturers and labs use to certify the tool, and they’re why you’ll hear techs talk about a wrench’s “full-scale” behavior.
According to research on ISO 6789 calibration, the standard simply doesn’t characterize that low end, so relative error can balloon on a job that sits near the bottom of a wrench’s scale. A fastener torqued near the bottom of a wrench’s range is operating in territory the certification may not have characterized.
The fix is straightforward:
- Choose a wrench whose working torque sits in the middle of its range, not near either end.
- For jobs that consistently fall below 20% of a tool’s scale, request calibration at that specific working point instead of relying on the standard 20/60/100 spread.
- Keep a second, lower-range wrench on hand for small fastener work rather than forcing one tool to cover everything.
Our sizing guide for shop torque wrenches walks through matching tool range to the fasteners you actually torque, which solves this problem before it starts.
Calibration Standards and How Often to Recalibrate
ISO 6789 splits into two parts that get confused constantly. Part 1 is a conformance declaration: it confirms a wrench meets its stated tolerance at the moment of testing, with no measurement uncertainty attached. Part 2 is a full calibration, and it’s the one that actually matters if you need traceable numbers. Standards research and a breakdown of ISO 6789:2017 calibration parameters both flag the same mistake: buying a Part 1 conformance check and treating it as a traceable calibration.
A real Part 2 certificate reports measured values at each test point, calculated uncertainty, and often repeatability, reproducibility, and length-dependency data, meaning how the reading changes with extensions or crow’s-foot adapters attached.
- Part 1 confirms conformance to a tolerance class with no uncertainty value.
- Part 2 delivers measured results, uncertainty, and direction-specific data.
- Length dependency testing shows how accessories shift your actual delivered torque.
For labs chasing the tightest possible numbers, NIST’s work on torque realization demonstrates calibration to 0.1% accuracy or better using realizers tied to fundamental constants, which is the benchmark reference labs use to calibrate the calibrators.
Manual torque wrenches are typically recalibrated every 12 months, per field verification guidance, with more frequent checks warranted for tools in heavy daily use or on safety-critical assemblies. Our own walkthrough on calibrating a torque wrench correctly covers when a shop check is enough and when it’s time to ship the tool out.
How Your Technique Changes the Torque You Actually Deliver
The wrench only does half the job. Your grip, pull speed, and hand position account for a real share of the error that shows up on a fastener, and most of it is avoidable.
- Keep your hand on the marked grip position, since the calibration assumes a specific effective length and sliding your palm up or down the handle changes the leverage.
- Exercise the mechanism with a few practice clicks before the job, because a wrench that’s been sitting can read differently until it’s cycled and settled.
- Pull smoothly and steadily rather than jerking the handle, since a fast yank can overshoot the click point before the mechanism registers it.
- Stop at the click on hard joints, but back off the instinct to keep cranking on soft or gasketed joints where overshoot happens easily.
- Account for extensions and crow’s-foot adapters, since they change the effective length and shift delivered torque unless you recalculate for the added reach.
- Match the wrench’s calibrated direction to your use, because a tool calibrated clockwise may not hold the same accuracy counterclockwise.
Pro Tip: Warm up a click wrench with two or three light clicks before the first real fastener. A mechanism that’s been idle in a tray for weeks needs to move before it clicks where it’s supposed to.
Our seven-step guide to calibrating click wrenches breaks each of these habits down further if you want the full shop routine.
Field-Testing Your Torque Wrench Between Lab Visits
You don’t need a lab to catch obvious problems between certifications. A few quick checks tell you whether a wrench is still trustworthy.
- Run a single-point check against a known weight or a second calibrated wrench at a torque value you use often.
- Test at your actual working point in the middle of the range, not just at the extremes, since that’s where you need the accuracy most.
- Check both directions if the job calls for clockwise and counterclockwise use, since the two can drift differently.
- Watch for a consistent bias, meaning the same tool reading high or low every time, which points to a calibration problem rather than random error.
Field checks are useful for catching drift, but they can’t replace a lab calibration when you need documented uncertainty. A lab tests multiple calibrated points and issues a certificate with that uncertainty attached, something a shop-floor comparison can’t produce. Call the lab after a drop, after storage under compression, or after an unusually heavy duty cycle, and anytime a field check shows a bias you can’t explain. Our quick shop-check guide covers the DIY side of this in more detail.
When ±4% Accuracy Is Good Enough for the Job
Not every fastener needs lab-grade precision. The real question is whether your joint’s tolerance leaves enough margin to absorb the tool’s rated error plus whatever bias your technique adds on top.
Picture a joint with a tight torque window. If the certificate shows the wrench running with a mean offset near the edge of its stated uncertainty, as the kind of result a Part 2 calibration would reveal, that offset alone can eat most of your available tolerance before you’ve even accounted for hand technique.
- Compare your joint’s tolerance, minus any process margin you need, against the tool’s rated uncertainty plus a reasonable allowance for operator variation.
- When the math leaves little room, request calibration at your working point instead of trusting a full-scale rating.
- For very low torque fasteners, a dedicated torque screwdriver beats forcing a general wrench near the bottom of its range.
- For high-torque applications beyond hand capacity, a torque multiplier keeps you in a controlled, repeatable range instead of guessing at applied force.
How We Build Torque Accuracy Into Our Tools
Our torque products are designed with ISO and DIN standards in mind. Our Get Organized™ foam trays help keep tools organized and reduce wrong-tool mix-ups.
Keeping tools in their marked spot also makes calibration tracking easier: you know at a glance which tool is due and which one just came back from the lab. Our guides on calibrating a torque wrench and choosing the right torque wrench walk through both sides of that process.
What Actually Matters Most for Accuracy
The accuracy number on the box gets more attention than it deserves. The standard tells you what the tool is capable of under lab conditions; it says nothing about what happens on your bench.
Conventional advice stops at “buy a good brand and calibrate once a year.” That’s a start, not a plan. The bigger lever is matching tool range to the job and treating technique as part of the accuracy spec rather than a separate concern. A perfectly calibrated wrench used near the bottom of its range, gripped at the wrong point, with a jerky pull, can miss a target by more than a cheaper tool used correctly in the middle of its range.
If you take one thing from all of this: read your joint’s actual tolerance before you worry about a tool’s accuracy class. Most fasteners have more margin than people assume, and a few genuinely don’t, which is exactly where sizing the tool and verifying it at the working point pays off.
— David
Pick the Right Torque Tools for Your Shop
Meeting a torque spec reliably starts with having the right tool for the range you’re actually working in, not one wrench stretched across every job. Our Torque Multiplier Set TTXMP12 and Torque Multiplier Set TEXMP12 give you controlled, repeatable high-torque delivery instead of guessing at applied force on large fasteners. For small, low-torque work, our Torque Screwdriver Set TTSD39 keeps you out of the bottom-range problem entirely.
Foam trays keep each piece in its own spot, so your calibrated tools stay organized and easy to track between uses. Browse our full torque wrench and tool lineup to find the right setup for your shop.
FAQ
Is a cheap torque wrench accurate?
Many budget torque wrenches meet their advertised accuracy, often around ±4% clockwise, when they’re new, according to independent testing. The bigger concern with inexpensive tools is unit-to-unit variability and faster drift over time compared to professional-grade wrenches.
How accurate are torque wrenches from Harbor Freight?
User testing and reviews found that Icon professional click-type wrenches from Harbor Freight are tested to ±4% clockwise and ±6% counterclockwise accuracy, performing competitively with pricier brands in head-to-head trials. Long-term consistency and unit-to-unit variation remain the main differences from premium tools.
How do I test a torque wrench for accuracy?
Run a single-point comparison against a known weight or a second calibrated wrench, ideally at the torque value you use most often, and check both directions if you work clockwise and counterclockwise. A consistent bias in your field check signals drift, but only a lab calibration with documented uncertainty confirms traceable accuracy.
How tight is 5 Nm by hand?
There’s no reliable way to judge a low torque by feel alone, which is exactly why low-torque fasteners call for a dedicated torque screwdriver rather than a hand-tightened guess. A tool sized for low-torque fasteners keeps the reading inside its characterized test points instead of relying on touch.
Sources
- How Accurate Are Harbor Freight Torque Wrenches? (According To Users)
- Leveraging the revised International System of Units: Torque from the fundamental constants | NIST
- Understanding torque calibration ISO 6789 standards
Key Standards, Labs, and Testing Articles
For further reading on the standards and testing behind this article, see NIST’s torque realization work, the ISO 6789 calibration breakdown, and independent wrench testing.