2021 test11 productsHand Tools

Which Torque Wrench Brand Wins?

We compared 11 torque wrench options head to head. Snap-on came out on top. See the measured results, the runner-up, the budget pick, and a link to the full test video.

The verdict
Winner

Snap-on

Price shown in test: $451

Check price on Amazon

Affiliate link. We may earn a commission at no extra cost to you.

The measured results

Every number below is read straight from the test. Scroll sideways to see all measurements. Products are listed in the order they finished.

ProductSpecsCw Test 50_initialCcw Test 250Cw 30 Measurement Aggregate_100to 250Cycle Test_1000cycles_at 50ftlbPhysical MeasurementsCcw 18 Measurement Aggregate_250_150_50
1Lexivon$4425 to 250 ft-lb range, reversible ratchet head (drives both directions but measures torque clockwise only), 2.5 ft-lb increments, 24-tooth chrome vanadium ratchet gear, claimed plus or minus 4% accuracy, 25 in length, made in Taiwan, includes carrying case and certificate of calibrationfirst two of six measurements within 4% at 50.85 and 51.46 ft-lb; last four overshot badly at 53.93, 55.46, 54.46, and 55.19 ft-lb; overall average (stated later in the cycle-test section) was 53.73 ft-lb before cyclingskipped, not designed for counterclockwise usenot named among the 3 brands that completed all 30 measurements within 4% (Icon, SK Tools, Snap-on), nor among the 3 named as outside the range just once (Kobalt, DeWalt, Proto); by elimination it had more than one out-of-range measurement across this test, but no specific count or average givenaveraged 53.73 ft-lb before 1,000 cycles; after cycling, average dropped to 52.01 ft-lb ('the spring steel has lost some of its strength'); 3 of 6 measurements fell outside the 4% range after cyclingno length, thickness, width, or swing-arc figure recorded for Lexivon in the video's physical-measurements comparison section beyond its intro length of 25 innot tested
2Performance Tool$76certified plus or minus 4% accuracy right-hand (clockwise) and 6% left-hand (counterclockwise), 25 to 250 ft-lb range, measures ft-lb and Newton meters, clockwise and counterclockwise operation, 41-tooth ratchet head, 22 in length, made in Taiwanovershot past the 50 ft-lb plus 4% range on all six of six measurements; third attempt was the worst at nearly 59 ft-lb; averaged 55.2 ft-lbovershot the first measurement by 28 ft-lb; outside the plus/minus range twice out of six measurements; averaged 261.8 ft-lbnot named among the completed-all-30 group or the outside-once group; by elimination had more than one out-of-range measurementaveraged 55.21 ft-lb before roughly 1,050 cycles; after cycling, average dropped nearly 5 ft-lb to 50.68 ft-lb; 2 of 6 measurements fell outside the 4% range after cycling22 in length noted in the intro; ratchet head width 1.63 in in the closing physical-measurements recapnot tested
3Craftsman current model$95certificate of calibration, 50 to 250 ft-lb range, inch and metric scale stamped into housing, plus or minus 4% clockwise / 6% counterclockwise accuracy, bi-material handle, locking feature, made in Taiwanwithin the 4% range on the first two attempts, overshot the next two, final two within range; averaged 52.18 ft-lb, off by 4.2%. Described as 'the best so far' at that point in the testing order (later brands performed better). Standard deviation for this test given later as 1.03 ft-lb.averaged 247.85 ft-lb, standard deviation 0.9 ft-lb, all six measurements within the 6% torque tolerance range, 'performed very well'not named among the completed-all-30 group or the outside-once group; by elimination had more than one out-of-range measurement. Standard deviation for this aggregate test given later as 1.12 ft-lb (a separate figure from the 1.03 ft-lb std dev at the initial 50 ft-lb test).performed nearly the same as before on the first three measurements, but overshot badly on the fourth and fifth; averaged 52.73 ft-lb after cyclinglength given as 27 and a quarter inches (27.25 in) in the intro, described at that point as 'by far the longest yet'; the closing recap instead states 'the Craftsman is nearly as long at 27.3,' a minor discrepancy from the intro figure, kept as stated rather than reconciled; front-to-back thickness 0.71 in (tied with DeWalt and Proto for thinnest); width 1.64 in (tied with DeWalt); swing arc 25 degrees (tied with DeWalt)one measurement fell outside the plus/minus 6% torque tolerance range across the 18-measurement counterclockwise test
4Kobalt$100certificate of calibration, 50 to 250 ft-lb range, clockwise only, positive locking mechanism, reversing ratcheting head (torques left and right hand fasteners though measurement is clockwise only), plus or minus 4% tolerance, SAE scale only (no metric), storage case described as 'not as durable as most of the other brands', 25 and 9/16 in length, made in Taiwanmuch more consistent than Lexivon, Performance Tool, and Craftsman, ranging from 51 to 52 ft-lb for an overall average of 51.25 ft-lb, standard deviation 0.42 ft-lb, 'a much tighter pattern than the previous three brands'doesn't advertise counterclockwise use but was tested anyway; performed better than Performance Tool but not as well as Craftsman, averaging 245.18 ft-lb; no explicit standard deviation given for Kobalt in this testoutside the plus/minus 4% range just one time (grouped with DeWalt and Proto)performed very well before 1,000 cycles, but struggled after with 4 of 6 measurements outside the plus/minus 4% range (3 too low, 1 too high), standard deviation over 2 ft-lb after cycling; no explicit before/after average figure given for Kobalt specifically in this section, only the qualitative struggle descriptionlength 25.6 in (tied with Snap-on for shortest of those named); swing arc 26.6 degrees, the widest (worst) of the swing-arc figures givennot tested
5Icon sold at Harbor Freight$12550 to 250 ft-lb range, certificate of calibration, 90-tooth ratcheting mechanism, professional standard accuracy plus or minus 4% clockwise / 6% counterclockwise, chrome plated hardened steel body, SAE and metric scale, roll mark scale for visibility, 23 and 9/16 in length, made in Taiwanperformed even better than Kobalt, most measurements within 1 ft-lb of the 50 ft-lb target; moved into the lead with an average accuracy of 50.7 ft-lb, 1.7% over target; standard deviation given later as 0.84 ft-lbaveraged 255.4 ft-lb, 2.1% over target, standard deviation 2.33 ft-lb. Transcript states 'the Icon was slightly more consistent than the Kobalt' immediately before this figure, though no comparable Kobalt standard deviation was given for this specific test to verify the comparison; kept verbatim.one of only 3 brands (with SK Tools and Snap-on) to complete all 30 measurements within the plus/minus 4% tolerance; came the closest to target torque on average of all brands, missing by only 0.2%; standard deviation for this aggregate test given later as 1.02 ft-lbaveraged 50.7 ft-lb before roughly 1,050 cycles with zero measurements outside the 4% range; did fairly well on the first five measurements after cycling but was off by nearly five and a half pounds on the sixth; standard deviation 2.45 ft-lb after cycling; post-cycle average given in the closing recap as 50.45 ft-lb (one of three brands, with SK Tools and Proto, that fell outside the 4% range only once after cycling)narrowest ratchet-head profile of all brands at 1.61 in width; swing arc 21 degrees, the second-best (tightest) of all brands after Snap-on's 15.6 degreesone measurement fell outside the plus/minus 6% torque tolerance range across the 18-measurement counterclockwise test (grouped with Craftsman)
6DeWalt$138certificate of calibration, 50 to 250 ft-lb range, protective head bumper, guaranteed accuracy 4% clockwise / 6% counterclockwise, locking trigger, bi-material grip, 1 ft-lb increments, SAE and metric scale, 27.5 in length (longest tool tested), made in Taiwanstarted just above 50 ft-lb on the first two measurements, just below 50 on the next three, then just above 50 again; averaged 50.11 ft-lb, standard deviation 0.38 ft-lb, 'moves into the lead over the Icon'. In the closing recap, DeWalt is named as coming the closest to 50 ft-lb on average of all brands, with all six measurements staying within the plus/minus 4% tolerance.averaged 1.4% under the 250 ft-lb target (versus 0.9% for Craftsman), but with a lower (better) standard deviation of only 0.67 ft-lb versus Craftsman's 0.9 ft-lboutside the plus/minus 4% range just one time (grouped with Kobalt and Proto)all six measurements within the plus/minus 4% range before roughly 1,050 cycles; after cycling, missed the target 2 of 6 times, overall average 51.33 ft-lb; standard deviation after cycling given as 1.6 ft-lb, the worst of the brands named in that recap (Snap-on 0.17, SK Tools 1.11, Wera 1.18, Proto 1.43, DeWalt 1.6)longest tool at 27.5 in; front-to-back thickness 0.71 in (tied with Craftsman and Proto for thinnest); width 1.64 in (tied with Craftsman); swing arc 25 degrees (tied with Craftsman)one of only 3 brands (with Snap-on and Proto) with zero measurements outside the plus/minus 6% tolerance across all 18 counterclockwise measurements; averaged 147.91 ft-lb specifically at the 150 ft-lb counterclockwise setting; standard deviation for this aggregate test given as 0.8 ft-lb (2nd best of the three, behind Proto's 0.6)
7SK Tools$200nice carrying case, certificate of calibration, locking collar handle, 50 to 250 ft-lb range, clockwise tolerance 4% / counterclockwise 6%, center-position ratchet lockout, dual SAE/metric scale, 24.25 in length, made in USA (US and foreign parts)a little high on the first measurement, a little low on the second, then a pretty tight pattern just above 50 on the final four; averaged 50.49 ft-lb, standard deviation 1.09 ft-lb, average within 1% accuracyovershot the target by 2.2% on average ('nearly as good as the Icon's 2.1%'), but with a better (lower) standard deviation of 1.45 ft-lb versus Icon's 2.33 ft-lbone of only 3 brands (with Icon and Snap-on) to complete all 30 measurements within the plus/minus 4% toleranceaveraged 50.49 ft-lb before roughly 1,050 cycles; after cycling, performed nearly the same at 50.17 ft-lb, though 1 of 6 measurements fell outside the 4% window; standard deviation after cycling was the second best of all brands at 1.11 ft-lb, behind only Snap-on's 0.17 ft-lbnot named in any of the four closing physical-measurement comparisons (thickness, width, length, swing arc); only its own intro-stated length of 24.25 in is availablenot tested
8Proto Proto Tools$236certificate of calibration, nice carrying case, marketed for 'precision highly accurate applications such as military and aerospace,' plus or minus 4% clockwise / 6% counterclockwise, SAE and metric scale, 26 and 5/16 in length, made in USAperformed well on the first attempt but overshot the second; attempts three through six were very close to 50 ft-lb; averaged 50.59 ft-lb, standard deviation 1.18 ft-lb; narrator's own comparative note: 'SK Tools did slightly better' (SK Tools std dev 1.09 at the same test)performed even better than SK Tools with an average of 2% over target, and a standard deviation of only 0.72 ft-lboutside the plus/minus 4% range just one time (grouped with Kobalt and DeWalt)averaged 50.59 ft-lb before roughly 1,050 cycles (same as the initial pre-cycle test average); after cycling, performed nearly the same at 50.24 ft-lb, though 1 of 6 measurements fell outside the 4% range; standard deviation after cycling was 1.43 ft-lbfront-to-back thickness 0.71 in (tied with Craftsman and DeWalt for thinnest); length 26.3 in in the closing recap (consistent with the intro's 26 and 5/16 in)one of only 3 brands (with Snap-on and DeWalt) with zero measurements outside the plus/minus 6% tolerance across all 18 counterclockwise measurements; averaged 152.7 ft-lb specifically at the 150 ft-lb counterclockwise setting; best standard deviation of the three at 0.6 ft-lb (versus DeWalt's 0.8 and Snap-on's 0.85)
9Wera$290certificate of calibration, 45 to 220 ft-lb range (narrower than most brands and does not officially cover the 250 ft-lb setting used elsewhere in testing), audible and tactile click, 45-tooth reversible ratchet mechanism, SAE and metric scale, clockwise only, 23 and 5/8 in length, made in Taiwanfirst four measurements within range, fifth was 'quite a bit under'; overall average 49.84 ft-lb, standard deviation 2.44 ft-lbskipped, not designed for counterclockwise usenot named among the completed-all-30 group or the outside-once group; by elimination had more than one out-of-range measurement. Note: Wera's stated torque range tops out at 220 ft-lb, yet the test protocol includes a 250 ft-lb setting; the transcript does not clarify how or whether Wera was tested at 250 ft-lb, flagged as an unresolved gap. Closing accuracy recap states Wera missed target torque by only 0.7% on average, the second-best result after Icon's 0.2%.averaged 49.84 ft-lb before roughly 1,050 cycles with 1 measurement outside the 4% window; after cycling, all six measurements were within the 4% window (improved), averaging 49.28 ft-lb; one of only 2 brands (with Snap-on) to have all six measurements within the 4% range after cycling; standard deviation after cycling was 1.18 ft-lbno thickness, width, or swing-arc figure recorded for Wera in the closing physical-measurements comparisons beyond its intro length of 23.625 innot tested
10Snap-on$451certificate of calibration, large carrying case ('seem to be larger than necessary'), clockwise only, 50 to 250 ft-lb range, SAE and metric scale, 25 and 9/16 in length, made in USAremained around 51 ft-lb, averaged 51.39 ft-lb (2.7% above the 50 ft-lb target), standard deviation 0.34 ft-lb, the best (most consistent) of all 10 brands at this test'did the best yet' at just 0.7% above target, standard deviation 0.52 ft-lb, best of all brands in this testone of only 3 brands (with Icon and SK Tools) to complete all 30 measurements within the plus/minus 4% tolerance; best standard deviation of all brands for this aggregate test at 0.51 ft-lbaveraged 51.38 ft-lb before roughly 1,050 cycles with standard deviation 0.34 ft-lb; after cycling, averaged 50.47 ft-lb with standard deviation improving to 0.17 ft-lb, the best of all 10 brands both before and after cycling; one of only 2 brands (with Wera) to keep all six measurements within the 4% range after cyclingthickness 0.75 in (second thinnest after the 0.71 in group); width 1.62 in (second narrowest after Icon's 1.61 in); length 25.6 in (tied with Kobalt); best (tightest) swing arc of all brands at 15.6 degreesone of only 3 brands (with DeWalt and Proto) with zero measurements outside the plus/minus 6% tolerance across all 18 counterclockwise measurements; came the closest to averaging 150 ft-lb of the three at 151.17 ft-lb, but had the worst (highest) standard deviation of the three at 0.85 ft-lb (versus Proto's 0.6 and DeWalt's 0.8)
11Craftsman old/vintage, approximately 11 years old at time of testing20 to 150 ft-lb range (narrower/lower than the current-model brands), clockwise only, made in USAgot off to a great start and things were going well until the last measurement, when it missed the target by nearly 4 ft-lb, for an overall average of 50.84 ft-lbnot testednot testednot testednot testednot tested

How it was tested

  • clockwise accuracy test at 50 ft-lb, six measurements per brand, using a Proto torque wrench tester accurate to 1/10 ft-lb
  • clockwise accuracy test at 100, 150, 200, and 250 ft-lb, six measurements each (30 total per brand)
  • counterclockwise accuracy test at 250, 150, and 50 ft-lb (18 total per brand where applicable; skipped for clockwise-only brands)
  • physical measurements (overall length, ratchet head thickness and width, swing arc with a socket attached to a fastener)
  • cycle durability test (each wrench cycled approximately 1,000 to 1,050 times, then retested for accuracy at 50 ft-lb)

if you want the best torque wrench, the Snap-on is definitely it. It just doesn't make mistakes.

From the test video verdict.

More Hand Tools