Key takeaways
- Nut-busting (Milwaukee, Makita), max breakaway (DeWalt) and max torque reverse (Chicago Pneumatic) are three names for the loosening figure. Fastening and max working torque forward are the tightening figure.
- None of the product pages or manuals checked here states how either figure is measured. The one test standard Chicago Pneumatic names in its CP7748 manual covers noise and vibration, not torque.
- The nut-busting to fastening ratio runs from 1.18 (Makita TW003G) to 1.73 (Chicago Pneumatic CP7748), with a median of 1.34 across nine wrenches. That spread is too wide to convert one figure into the other.
- The battery is part of the rating. Milwaukee's 2967 gains 100 ft-lb on both figures when the XC5.0 pack is swapped for a FORGE pack, and DeWalt ties its 1,320 ft-lb to one named pack.
- No figure on the box means the wrench tightens to a spec. Chicago Pneumatic's lug nut version stops at 66 to 71 ft-lb and still says final torque goes on with a calibrated torque wrench.
Two readings of the bigger number circulate. One says nut-busting is just marketing, the same torque measured more generously. The other treats it as the wrench's real strength and fastening as a cautious footnote. Neither is right. The two figures describe two different jobs, loosening in reverse and tightening in forward, and the makers keep them apart. What they do not do is say how either is measured.
What the makers publish, and what they leave out
Every number on this page comes from a maker's product page, manual or leaflet, or a retailer listing that quotes the maker's rating. None of those documents, from any of the four makers, states a test method for either figure: no joint, no fixture, no duration, no standard.
The closest any of them comes is instructive. Chicago Pneumatic's CP7748 manual names two ISO standards, ISO 15744 and ISO 28927-2, and says the declared values "were obtained by laboratory type testing in accordance with the stated standards and are suitable for comparison with the declared values of other tools tested in accordance with the same standards." Those standards cover the noise and vibration columns. The torque columns beside them carry no such statement.
So the comparison below is built on what every maker does publish: a pair of numbers, one per direction, on one tool, under that maker's own conditions. The ratio between them is the fairest thing to compare across brands.
Three names for the loosening figure
The same idea goes by three labels:
| Maker | Tightening figure, as printed | Loosening figure, as printed |
|---|---|---|
| Milwaukee | Fastening torque | Nut-busting torque |
| Makita | Max fastening torque | Nut-busting torque |
| DeWalt | Max fastening torque | Max BREAKAWAY torque |
| Chicago Pneumatic | Max working torque fwd | Max torque reverse |
Chicago Pneumatic's labels are the plainest, because they name the direction. Its CP7748 manual goes one step further and heads the two torque columns Working and Max. The forward figure is printed as a range, 53 to 575 ft-lb (72 to 779 Nm), and the other as a single maximum, 922 ft-lb (1,250 Nm). The product page gives 553 and 959 ft-lb; this page uses the product pages throughout, because they give all three CP models in one format.
The words matter for reading a cordless spec sheet. "Fastening" and "working forward" describe what the tool is rated to do in the direction you tighten. "Nut-busting", "breakaway" and "reverse" describe the direction you loosen. A wrench with 1,600 ft-lb of nut-busting torque is not rated to tighten anything to 1,600 ft-lb.
Nine wrenches, both figures, one ratio
Every published pair checked here, with both units and the battery each figure is stated on:
| Tool | Power, as stated | Fastening | Nut-busting | Ratio |
|---|---|---|---|---|
| Makita TW003G (1/2 in) | 40V max XGT | 1,003 ft-lb (1,360 Nm) | 1,180 ft-lb (1,600 Nm) | 1.18 |
| Makita TW002G (1/2 in) | 40V max XGT | 1,254 ft-lb (1,700 Nm) | 1,623 ft-lb (2,200 Nm) | 1.29 |
| Chicago Pneumatic CP7782 (1 in) | Air, 90 psi max | 1,480 ft-lb (2,007 Nm) | 1,920 ft-lb (2,603 Nm) | 1.30 |
| Milwaukee 2967 (1/2 in) | M18 FORGE pack | 1,200 ft-lb (1,627 Nm) | 1,600 ft-lb (2,169 Nm) | 1.33 |
| Chicago Pneumatic CP7737 (1/2 in) | Air, 90 psi max | 165 ft-lb (224 Nm) | 221 ft-lb (300 Nm) | 1.34 |
| Milwaukee 2967 (1/2 in) | M18 XC5.0 pack | 1,100 ft-lb (1,491 Nm) | 1,500 ft-lb (2,034 Nm) | 1.36 |
| DeWalt DCF900 (1/2 in) | 20V MAX, no pack named | 1,030 ft-lb (1,396 Nm) | 1,400 ft-lb (1,898 Nm) | 1.36 |
| DeWalt DCF961 (1/2 in) | 20V MAX, fastening on DCBP520 | 1,320 ft-lb (1,790 Nm) | 1,900 ft-lb (2,576 Nm) | 1.44 |
| Chicago Pneumatic CP7748 (1/2 in) | Air, 90 psi max | 553 ft-lb (750 Nm) | 959 ft-lb (1,300 Nm) | 1.73 |
The median of the nine ratios is 1.34. The middle of the table is tight: six of the nine sit between 1.29 and 1.36. The ends are not. Makita's TW003G puts its loosening figure only 18 percent above its tightening figure. Chicago Pneumatic's CP7748 puts it 73 percent above, and that is one maker, one 1/2 in tool, with both figures on the same page.
That spread is the practical finding. If a wrench publishes only one number, there is no reliable factor to recover the other. Divide a 1,500 ft-lb nut-busting figure by the ends of this range and the tightening rating could be anywhere from about 870 to 1,270 ft-lb. Read the figure that matches the job, and if the maker does not print it, treat the tool as unrated for that job.
Why the bigger number is not the stronger tool
Put Chicago Pneumatic's CP7748 beside Milwaukee's 2967 on the XC5.0 pack. By the loosening figures, 959 against 1,500 ft-lb, the cordless tool rates about 56 percent higher. By the tightening figures, 553 against 1,100 ft-lb, it rates about twice as high. Which gap matters depends on the job, and neither controls for the test.
Now mix the two by mistake, as a casual spec comparison often does: the CP7748's 959 ft-lb reverse figure against the 2967's 1,100 ft-lb fastening figure. The tools look within 15 percent of each other. On matched figures they are 56 to 99 percent apart. A table that lines up a "max torque" column without saying which direction each entry is has already made that mistake.
Air supply matters for the pneumatic rows. Chicago Pneumatic lists 90 psi as the maximum dynamic working pressure for all three of its wrenches here. The air hose size guide works out that 50 ft of 3/8 in hose loses about 30 psi at the CP7748's 25 CFM continuous draw, and the compressor tank guide shows how quickly a small pump falls behind that flow. A hose that drops 30 psi leaves the wrench working well below the pressure its page lists, whichever direction it is turning.
The battery is part of the rating
Milwaukee's two published pairs for the 2967 are the clearest evidence that a cordless torque figure belongs to a tool and a pack together:
| 2967 paired with | Fastening | Nut-busting | Ratio |
|---|---|---|---|
| M18 REDLITHIUM XC5.0 | 1,100 ft-lb | 1,500 ft-lb | 1.36 |
| M18 REDLITHIUM FORGE | 1,200 ft-lb | 1,600 ft-lb | 1.33 |
| Difference | +100 ft-lb (+9.1%) | +100 ft-lb (+6.7%) |
Same tool, same anvil, same modes; a different pack adds 100 ft-lb to both figures. DeWalt states the same dependency in fewer words. Its DCF961 page gives the 1,320 ft-lb fastening figure "Based on using DCBP520 (sold separately)". The page attaches no battery to the 1,900 ft-lb breakaway figure. Makita's TW003G page lists five compatible XGT pack sizes, from 2Ah to 8Ah, and does not say which one its 1,360 and 1,600 Nm were reached on.
Makita's own TW0350 manual names supply as a factor outright: "Voltage drop will cause a reduction in the fastening torque." So a published cordless figure is the tool on the pack the maker chose. On a smaller, older or cold pack, expect less; the charge and temperature limits the three cordless makers print are collected in whether tool batteries can stay in a truck box. When you compare packs across brands, use watt hours rather than the voltage on the label, the arithmetic worked through in whether 20V MAX is the same as 18V.
Time is part of it too
DeWalt's DCF961 page carries one more condition that changes how the breakaway number reads: "After 4 seconds of impacting, the tool automatically provides an increase in power to aid in loosening stubborn fasteners." The page does not say whether the 1,900 ft-lb figure includes that boost. It does establish that the tool's output in reverse is not one fixed value from the first blow.
Makita's manuals treat tightening the same way. The TW0350 manual says "the relation between fastening torque and fastening time is shown in Fig. 4 and Fig. 5" and tells the user to "fasten for the proper fastening time." It warns that fastening longer than the figures show can overstress the bolt or socket, and that a test operation should be run first to set the time for each bolt. In Makita's own description, the torque a fastener ends up with is something you approach by holding the trigger, not a value the tool delivers and stops at.
So a fastening rating is a ceiling on the maker's joint in the maker's time. What your joint gets depends on factors Makita lists in the same manual:
- Voltage. A drop reduces fastening torque.
- Socket. The wrong size, or a worn hex or square end, reduces it.
- Bolt. Diameter, class, torque coefficient and length all change the proper fastening torque.
- Extensions and universal joints. These "somewhat" reduce the fastening force; Makita says to compensate by fastening longer.
- How the tool is held and the material at the fastening position.
Impact wrenches do not torque lug nuts to spec
A wrench rated 1,100 ft-lb fastening has a mode that stops early, so it seems a short step to setting lug nuts with it. The makers' own documents say otherwise.
Chicago Pneumatic builds a torque-limited version of the CP7748 for exactly this job. Its manual describes the CP7748TL as "a Torque Limited Impact Wrench intended for initial tightening of wheel lug nuts and bolts", rates its working torque at 66 to 71 ft-lb (90 to 96 Nm), and then instructs: "The operator should release the trigger when impacting begins, and the fastener is snug. Final torque should be applied manually using a calibrated torque wrench, and following the vehicle manufacturer's instructions." The limited tool snugs; the torque wrench finishes.
Makita's TW0350 manual is shorter: "After fastening, always check the torque with a torque wrench." Chicago Pneumatic's own guide to choosing between tools says impact wrenches are "mainly used for dismantling nuts as they do not deliver accuracy for bolting."
The low-torque modes on cordless wrenches point the same way. Milwaukee describes the 2967's auto shut-off as a "precise 50 ft-lbs torque limit". DeWalt says its Precision Wrench mode prevents overtightening and fastener run-off. Both are there to stop the tool going too far. Neither is a calibrated final torque.
So the sequence for wheels is: run the nuts down with the impact in a limited mode, stop at snug, then bring each nut to the vehicle maker's figure with a click torque wrench, following the vehicle manufacturer's instructions as Chicago Pneumatic puts it. That wrench is a calibrated instrument, and how you store it decides whether its number holds. The torque wrench storage guide covers why it goes back in the drawer at the lowest mark on its scale, not at zero, and how often it needs checking.
Reading a spec sheet before you buy
The figures are least misleading when read in this order:
- Find both numbers and their labels. If the sheet gives one "max torque" with no direction, do not guess which it is. Check the maker's own page for a fastening or forward working figure before you compare.
- Note the power source each figure is stated on. For cordless tools, that means the named pack. For air tools, note the working pressure the page lists, 90 psi on Chicago Pneumatic's.
- Match the figure to the job. Removing rusted suspension and axle nuts is a loosening job; compare nut-busting figures. Running down large bolts on equipment is a tightening job; compare fastening figures.
- Do your own unit conversion. Makita's Czech leaflet for the XGT wrenches prints the TW003G as "1,360 N·m (1,250 ft.lbs.)" fastening and "1,600 N·m (1,620 ft.lbs.)" nut-busting. Those imperial values do not match the metric ones. At 0.7376, the correct figures are about 1,003 and 1,180 ft-lb, and the printed ones match the TW002G column beside it.
- Expect a ratio near 1.3, and read twice when it is far off. Six of nine pairs here fall between 1.29 and 1.36; the CP7748's 1.73 is correct as printed, but unusual.
A high-torque 1/2 in impact also changes what goes in the drawer beside it. The 2967 and the TW002G use a friction ring anvil and the TW003G a detent pin, so check which your sockets suit; the socket organizer guide covers rail and tray layouts for a deep impact set.
Where this answer stops applying
- Other makers and models. The ratio range is drawn from nine published pairs. It describes these tools, not every impact wrench, and a maker not listed here may define its figures differently.
- Any claim about the test. None of the documents checked here states how either figure is measured, so this page does not say what joint, duration or fixture produced any of them. The comparison is between published numbers, not between measured tools.
- Your joint. A rated figure is a ceiling on the maker's joint. Voltage, socket fit, bolt size and class, extensions and air pressure all move what a given fastener actually gets.