Skip to content
Pulse and Path

The evidence scoreboard

The best fitness trackers for walking

The published error range for step counting runs from -9% to 12% across the whole category. That number is why this page ranks on everything except the step count.

By Stephen V.Published November 22, 2026
#ad

We earn a commission when you buy through our Amazon links, at no extra cost to you. It never decides a ranking — the rubric that does is published in full, so you can check us rather than take our word for it. How this works.

Fitbit Fitbit Charge 6

Top pick

Lowest-cost tracker with published data behind it

Fitbit Charge 6

Cheap, measured, and the weakest of the three devices in the PSG study. We would rather tell you that than not rank it.

Peer-reviewed validation7.2/10
Check price on Amazon →

No buyable offer at the last price check (November 21, 2026). We show nothing rather than a stale number.

#ad · we may earn a commission from this link to Fitbit Charge 6

The short answer

The top pick and the budget pick are the same device: the Fitbit Charge 6, the cheapest tracker on this roster with a published validation study behind its platform. If you want satellite accuracy and a watch rather than a band, the Garmin vivoactive 6 states 21 hours of GPS and pairs a chest strap. For walking specifically, do not pay for multi-band positioning - nothing here has it, and you do not need it.

Ranked

Quick picks

#DeviceBest forEvidenceScorePrice
01
Fitbit Fitbit Charge 6
Fitbit Charge 6

Cheap, measured, and the weakest of the three devices in the PSG study. We would rather tell you that than not rank it.

Lowest-cost tracker with published data behind itPeer-reviewed validation7.2/10Check price →
02
Garmin Garmin vívoactive 6
Garmin vívoactive 6

The cheapest Garmin that still pairs a chest strap and still estimates VO2 max — and the published battery table shows you why it outlasts a Forerunner.

A $300 Garmin for someone who already owns the strapManufacturer specification6.6/10$249.00Amazon
03
Amazfit Amazfit Balance 2
Amazfit Balance 2

The most published capability per dollar on this site, and the least checkable. Dual-band GNSS, offline maps and sapphire glass, from a brand no validation study we could find has ever included.

Spec-sheet value, if unverified data is a trade you acceptManufacturer claim only6.5/10$199.99Amazon
04
Garmin Garmin Forerunner 165
Garmin Forerunner 165

The cheapest way into a genuinely good AMOLED Garmin, provided you accept single-band positioning.

First real running watchManufacturer specification6.4/10$191.00Amazon

#ad· Live prices from the Amazon Product API, as of November 21, 2026. Where we have no verified live price we show none. Scores come from the published rubric at How we rank — the Evidence column shows what each ranking actually rests on, including where the answer is nothing.

The metric that does not separate anything

Walking means step count, and step count is the one thing this category is broadly good at. The 2024 living umbrella review - 24 systematic reviews, 249 primary validation studies, 430,465 participants - reports step-count mean absolute percentage error from -9% to 12% across the field, alongside heart rate at roughly plus or minus 3%. That range is wide in absolute terms and narrow in the way that matters here: it is the range for consumer wearables generally, not a ranking that puts one of these four above another.

The same review found roughly 34 of 310 identified wearables had ever been validated at all. So a page that ranked walking trackers by step accuracy would be ranking on a number nobody has published per device. This one ranks on four things that arepublished, and says plainly that the step count is not one of them.

  • Stated battery. A walking tracker is worn all day and all night. The charger is the real enemy.
  • Its own satellite receiver, or your phone's. Distance and route without the phone in your pocket is the single biggest capability difference in this field.
  • Whether it will read an external sensor. Walking heart rate is low and steady, which is the easiest case for a wrist sensor - but the option still matters if you ever do anything else.
  • Price.Three of these four are under $250 at today's prices.

1. Fitbit Charge 6 - the cheapest device here with a study behind it

The Charge 6 scores 7.2 and is the only pick on this page whose platform has been tested against a gold standard. In Robbins et al. (2024), 35 adults slept an inpatient night with polysomnography wearing three consumer devices, and Fitbit was one of them - against a category where almost nothing has been tested at all.

Now the unflattering half, in the same breath, because it is the same study: the Fitbit device tested came third of three, at 70.9% four-stage agreement and 61.7% deep-sleep sensitivity, overestimating light sleep by 18 minutes and underestimating deep sleep by 15. And the device measured was the Sense 2, not the Charge 6 - Fitbit shares its staging algorithm across the line, but that is still evidence about a sibling.

For a walker it is still the right first pick: a slim band rather than a watch, up to 7 stated days, built-in GPS so distance does not depend on your phone, heart rate broadcast to connected gym equipment, and (no verified live price today, so we are not calling this one either way). Some features sit behind Fitbit Premium, which is the cost you should check before buying.

2. Garmin vivoactive 6 - the longest satellite day, and it pairs a strap

Garmin's manual states up to 11 days in smartwatch mode, 21 days in battery saver, 21 hours GPS only, 17 hours all satellite systems and 8 hours all-systems with music. Twenty-one stated hours of GPS is far more walking than any single day contains, and the 11-day smartwatch figure is the practical one.

The manual also documents ANT+ and Bluetooth sensor pairing, naming the HRM 200, HRM 600, HRM-Fit and HRM-Pro series - the one real fix for wrist heart rate, on Garmin's cheapest current AMOLED watch. It scores 6.6 and under $500 today. Note the shape of that battery table: there is no all-systems-plus-multi-band row, because the hardware has no multi-band mode. For walking that absence costs you nothing; for city running it would.

3. Amazfit Balance 2 - the longest stated battery, and the weakest evidence

Amazfit states a 47.4 mm case at 12.3 mm thick excluding the sensor, sapphire glass, a 1.5-inch AMOLED at a stated 2,000 nits peak, 10 ATM water resistance, a 658 mAh battery rated at "up to 21 days" of typical use and up to 33 hours in Accuracy GPS Mode, dual-band support from six satellite systems, and free downloadable offline maps with turn-by-turn directions.

Twenty-one stated days is the longest figure on this page by a wide margin, and offline maps on a walking watch are genuinely useful. It still ranks third, at 6.5, because of the 30% evidence weight: Amazfit's own published answer to the accuracy question is "industry-leading GPS accuracy" with no method, no instrument and no figure. Three absences are recorded from its own page as absences rather than as claims: it does not name VO2 max, does not state whether an external chest strap can be paired, and names only the Zepp app as a destination for your data.

4. Garmin Forerunner 165 - the running watch that walks well

Garmin's manual states up to 11 days in smartwatch mode with activity tracking and wrist heart rate, 19 hours GPS Only, 17 hours All Systems, 25 hours UltraTrac, 7 hours GPS Only with music, a 5 ATM swim rating and up to 4 GB of media storage. It scores 6.4 and under $200 today.

It is on this page because of UltraTrac: 25 stated hours of position logging is the longest single-session satellite figure here, which suits a long day on foot better than anything else at this price. Don't buy it if you want a lifestyle watch - there are no maps, there is no multi-band row in that table because the hardware has no multi-band mode, and the interface is built for running.

What not to buy for walking

Two things, stated plainly so you keep the money. Do not pay for multi-band positioning. None of these four has it, and the error it fixes - signal bouncing off buildings and canyon walls at speed - barely applies to walking pace. Do not buy a ring for step counting.No published study has ever measured a ring's step count, and the structural reason is simple: a hand swings, gestures and rests on things independently of your legs. Rings are excellent at sleep and have no step-counting evidence at all, which is why none appears above.

In full

Every pick, and why

01Lowest-cost tracker with published data behind it

Fitbit Charge 6

Cheap, measured, and the weakest of the three devices in the PSG study. We would rather tell you that than not rank it.

Form
Wrist band
Battery (stated)
Up to 7 days
Extras
Heart rate on connected gym equipment

What's good

  • One of only three devices with a published head-to-head PSG result at all
  • Substantially cheaper than every ring here

What isn't

  • Robbins et al. (2024): lowest four-stage agreement of the three tested, at 70.9%, and deep-sleep sensitivity of 61.7%
  • Overestimated light sleep by 18 minutes and underestimated deep sleep by 15

Skip this if

Skip this if the sleep stages are the point. It was measured, and it came third of three.

What the evidence says

Peer-reviewed validation

Robbins et al., Sensors 24(20):6532 (2024) tested the Fitbit Sense 2 — not the Charge 6 — against polysomnography: 70.9% four-stage agreement (κ = 0.55) and 61.7% deep-sleep sensitivity, the weakest of the three devices. Fitbit's staging algorithm is shared across the line, but this remains evidence about a different device.

Show how this 7.2 was calculated
FactorWeightScore
EvidenceReal peer-reviewed evidence exists for the platform — it is simply unflattering, and it tested a sibling device.30%6.0
HardwareWrist-worn optical sensing, the weakest of the three placements in the placement study.20%7.0
Metric coverageBroad everyday metrics; some features sit behind Fitbit Premium.20%7.5
EnduranceUp to 7 stated days.15%7.5
Cost per capabilityBy far the cheapest device in this roundup.15%9.0
Fitbit Fitbit Charge 6
Our score7.2/10
Check price on Amazon →

No buyable offer at the last price check (November 21, 2026). We show nothing rather than a stale number.

#ad · we may earn a commission from this link to Fitbit Charge 6

02A $300 Garmin for someone who already owns the strap

Garmin vívoactive 6

The cheapest Garmin that still pairs a chest strap and still estimates VO2 max — and the published battery table shows you why it outlasts a Forerunner.

Case
42mm, aluminum bezel, AMOLED
Satellite
GPS only or all systems — no multi-band row published
Battery (stated)
11 days smartwatch / 17 h all satellite systems
Maps
None
Sensors
Pairs ANT+ / Bluetooth straps, HRM 600 and HRM 200 named

What's good

  • Garmin's manual documents ANT+ and Bluetooth strap pairing, so the one real fix for wrist heart rate works on a $300 watch
  • Carries its own VO2 max topic in the manual, plus PacePro, running dynamics, HRV status and Body Battery
  • States longer endurance than a Forerunner 570 in every satellite mode it publishes — 21 h GPS-only against 18 h

What isn't

  • No multi-band row exists anywhere in Garmin's own battery table, which is how you can tell the receiver is single-frequency
  • Garmin's feature list names no training status, no acute load and no training readiness — the analysis layer a Forerunner or a Venu 4 gets
  • No maps, one case size only, and no independent validation of anything it measures

Skip this if

Skip this if positioning accuracy is the weakness you are upgrading. The long battery here is partly a consequence of hardware that cannot run a multi-band mode, and no amount of software fixes a single-frequency receiver in a city or a canyon.

What the evidence says

Manufacturer specification

No independent validation of this model exists that we could find — not of its optical heart rate, not of its sleep staging, not of its positioning. Garmin as a brand has at least been measured: two Garmin watches were among the eight tested against a trundle wheel by Gilgen-Ammann et al. (2020), and a Garmin device finished last of five on nocturnal HRV agreement in Dial et al. (2025). Neither study tested this watch, and the 2020 one predates the receiver generation it would need to describe.

Show how this 6.6 was calculated
FactorWeightScore
EvidenceA detailed owner's manual with a per-mode battery table and documented feature definitions, and no independent measurement of a single number this watch produces.30%4.5
HardwareGarmin's Elevate wrist sensor, an AMOLED display behind an aluminum bezel, multi-constellation GNSS — and no multi-band mode, which Garmin's own battery table gives away by having no row for one. Pairs ANT+ and Bluetooth sensors, including the HRM 600, which is worth more to heart-rate data than any wrist sensor revision.20%7.0
Metric coverageVO2 max, PacePro, running dynamics, HRV status, Body Battery, sleep coach, Pulse Ox, stress, 80+ sports apps and a wheelchair mode. Missing: maps, an ECG app, and the training status / acute load / training readiness layer Garmin names on the tier above.20%7.5
EnduranceGarmin states 11 smartwatch days, 21 days in battery saver, 21 hours GPS-only and 17 hours on all satellite systems. Longer than a Forerunner 570 in every mode both watches publish.15%8.0
Cost per capabilityA $299.99 list price for a Garmin that estimates VO2 max and pairs a strap is a genuinely good number. It is not a great one, because an Amazfit Balance 2 publishes dual-band GNSS, offline maps, sapphire glass and 10 ATM at the same list price.15%7.5
Garmin Garmin vívoactive 6
Our score6.6/10
$249.00Check price on Amazon

$299.9917% off

Price as of November 21, 2026. Prices move; Amazon's is the one that counts.

#ad · we may earn a commission from this link to Garmin vívoactive 6

03Spec-sheet value, if unverified data is a trade you accept

Amazfit Balance 2

The most published capability per dollar on this site, and the least checkable. Dual-band GNSS, offline maps and sapphire glass, from a brand no validation study we could find has ever included.

Case
47.4mm x 12.3mm, aluminum frame, sapphire lens
Satellite
Dual-band, six satellite systems (manufacturer-stated)
Battery (stated)
21 days typical / 33 h Accuracy GPS Mode
Maps
Free offline maps, turn-by-turn, offline routing
Water
10 ATM (manufacturer-stated)

What's good

  • Publishes dual-band GNSS, free offline maps, sapphire glass and 10 ATM at a list price where Garmin publishes none of the four
  • 33 stated hours in its dual-band Accuracy GPS Mode — roughly double what a similarly priced watch here states in its own best-accuracy mode
  • No subscription named anywhere on the page for workout and recovery data, which is not true of every rival in this category

What isn't

  • Near the bottom of our evidence hierarchy, and worse than empty: Amazfit answers the accuracy question with the phrase "industry-leading GPS accuracy" and publishes no method, no reference instrument and no figure
  • Amazfit's own specification page does not name VO2 max, does not state whether a chest strap can be paired, and names no companion app other than Zepp
  • Nobody outside the company appears to have measured any Amazfit against a reference in anything we were able to read

Skip this if

Skip this if the reason you are upgrading is that you stopped trusting your current watch's numbers. Buying a better spec sheet from a brand with no measurement record does not fix that problem, it relocates it — and if a chest strap is part of your plan, Amazfit's page does not say this watch will talk to one.

What the evidence says

Manufacturer claim only

No peer-reviewed validation, no independent measured test and no published method of any kind for this watch, and we found no validation of any Amazfit device in anything we were able to read — the brand was not among the eight watches in Gilgen-Ammann et al. (2020), not in Robbins et al. (2024) and not in Dial et al. (2025). Amazfit does publish an accuracy claim: its own FAQ answers "how accurate is the GPS" with "industry-leading GPS accuracy". That is a manufacturer claim with no method behind it, which is the second-weakest tier we have.

Show how this 6.5 was calculated
FactorWeightScore
EvidenceHalf a point BELOW the floor we give a maker with simply no measurement record, because Amazfit does not merely lack a validation — it publishes an accuracy claim with no method, no reference and no number attached, which is the specific failure this site exists to flag.30%2.5
HardwareDual-band reception across six stated constellations, a sapphire lens, a 658 mAh cell, 10 ATM and a 2,000-nit 1.5-inch AMOLED. On published hardware this is the strongest specification at its price anywhere on this site.20%8.5
Metric coverageTraining Load, Recovery Status, HybridCharge, Zepp Coach, sleep HRV, blood oxygen, stress, 180+ sports modes and genuinely free offline maps. Marked down for what the page does not say: no VO2 max, no statement that an external heart-rate strap can be paired, and no route out of the Zepp app described anywhere on it.20%6.5
Endurance21 stated days of typical use and 33 stated hours in the dual-band Accuracy GPS Mode — the best stated endurance per dollar on this site, and stated in the accuracy mode rather than a flattering power-saving one.15%9.0
Cost per capabilityLists at $299.99, sells below it, and delivers dual-band GNSS plus offline maps for less than half what the cheapest Garmin with both costs. On arithmetic alone this is the best cost-per-published-capability figure here.15%9.0
Amazfit Amazfit Balance 2
Our score6.5/10
$199.99Check price on Amazon

$219.999% off

Price as of November 21, 2026. Prices move; Amazon's is the one that counts.

#ad · we may earn a commission from this link to Amazfit Balance 2

04First real running watch

Garmin Forerunner 165

The cheapest way into a genuinely good AMOLED Garmin, provided you accept single-band positioning.

Display
AMOLED
Satellite
Single-band GNSS
Extras
Training metrics, sleep tracking, music on the Music edition

What's good

  • AMOLED and Garmin's training platform at the bottom of the range
  • Light, simple, and hard to get wrong as a first watch

What isn't

  • Single-band positioning, which is the specification that actually degrades in cities and under tree cover
  • No multi-band means it is the wrong upgrade if GPS drift is your complaint

Skip this if

Skip this if you are here because your GPS track wanders between buildings. Single-band is the hardware that causes that, and this watch has it.

What the evidence says

Manufacturer specification

Published specifications only; no independent validation of this model that we could find.

Show how this 6.4 was calculated
FactorWeightScore
EvidencePublished specs, no independent measured validation.30%5.0
HardwareSingle-band GNSS — a genuine step down from the multi-band models above it.20%6.0
Metric coverageFull Garmin training platform; fewer advanced running-dynamics metrics.20%7.0
EnduranceSolid for the class, per Garmin's stated figures.15%7.0
Cost per capabilityLowest entry price to the Garmin ecosystem with an AMOLED screen.15%8.5
Garmin Garmin Forerunner 165
Our score6.4/10
$191.00Check price on Amazon

$249.9924% off

Price as of November 21, 2026. Prices move; Amazon's is the one that counts.

#ad · we may earn a commission from this link to Garmin Forerunner 165

How this ranking was built

Every score on this page comes from one published rubric, applied identically to every device on the site, with the weights written down and the inputs shown. We have not worn any of these devices and we never say we have. Read the method and check us.

Show your working

Sources

Every one of these was read in full before it was cited. Where we could not access a paper, we do not quote its numbers.

  1. [1]Peer-reviewed validation2024
    Doherty, Baldwin, Keogh, Caulfield & Argent, “Keeping Pace with Wearables: A Living Umbrella Review of Systematic Reviews Evaluating the Accuracy of Consumer Wearable Technologies in Health Measurement”, Sports Medicine

    DOI 10.1007/s40279-024-02077-2. An umbrella review: 24 systematic reviews covering 249 unique primary validation studies and 430,465 participants (43% female), across 11 biometric outcomes, with roughly 34 of 310 identified consumer wearables ever validated. Reported accuracy by metric: heart rate approximately ±3% (mean bias −3.39 bpm); step count mean absolute percentage error from −9% to 12%; energy expenditure mean bias −3 kcal per minute with error ranging from −21.27% to 14.76%. Authors’ stated limits: a temporal lag between device release and published research, the intrinsic bias of a tertiary synthesis, and heterogeneity of protocols, devices, criterion measures and statistics severe enough that the authors did not pool a meta-analysis.

  2. [2]Peer-reviewed validation2024
    Robbins et al., “Accuracy of Three Commercial Wearable Devices for Sleep Tracking in Healthy Adults”, Sensors 24(20):6532

    Brigham and Women’s Hospital. n = 35 healthy adults, single-night inpatient polysomnography, testing Oura Ring Gen3, Fitbit Sense 2 and Apple Watch Series 8. Authors’ stated limits: one night, lab conditions, healthy adults without insomnia, comparison only during scheduled sleep, and device failures on 6 Apple Watches and 2 Fitbits.

  3. [3]Manufacturer specification2026
    Garmin — vívoactive 6 Owner’s Manual, battery life information

    Manufacturer-stated, read 2026-11-01 from Garmin's own owner's manual: smartwatch mode up to 11 days; battery saver watch mode up to 21 days; GPS only mode up to 21 hours; all satellite systems mode up to 17 hours; all satellite systems with music mode up to 8 hours. Garmin's own caveat on the page: “the actual battery life depends on the features enabled on your watch.” NOTE the structure of the table — the highest-accuracy row Garmin publishes for this watch is all satellite systems. There is no all-systems-plus-multi-band row, because the hardware has no multi-band mode.

  4. [4]Manufacturer specification2026
    Garmin — vívoactive 6 Owner’s Manual, “Pairing Your Wireless Sensors”

    Manufacturer-documented, read 2026-11-01. The watch pairs wireless sensors “using ANT+ or Bluetooth technology” and reconnects automatically “when you start an activity and the sensor is active and within range”. The US manual names the HRM 200, HRM 600, HRM-Fit and HRM-Pro series, and advises pairing an HRM-Fit or HRM-Pro over ANT+ rather than Bluetooth when you want running pace and distance from the strap. This is the single most consequential line in the manual for anyone who cares about heart-rate data: the watch can be told to stop guessing and read a chest electrode instead.

  5. [5]Manufacturer specification2026
    Amazfit — Balance 2 product page and specification table (US)

    Manufacturer-stated, read 2026-11-01. Case 47.4 mm diameter and 12.3 mm thick excluding the heart-rate sensor, aluminum alloy frame with a fiber-reinforced polymer case, scratch-resistant sapphire glass lens, 1.5-inch AMOLED at a stated 2,000 nits peak brightness, 10 ATM water resistance, a 658 mAh battery stated at “up to 21 days” of typical use and up to 33 hours in Accuracy GPS Mode, “dual-band support from six satellite systems”, free downloadable offline maps with turn-by-turn directions and offline route planning, 180+ sports modes, BioTracker monitoring of heart rate, sleep, blood oxygen and stress plus sleep HRV, Training Load, Recovery Status, HybridCharge and Zepp Coach, and “no subscription workout and recovery data in the Zepp App”. List price $299.99. THREE ABSENCES, recorded as absences rather than as claims about what the watch cannot do: the page does not name VO2 max, does not state whether external Bluetooth or ANT+ sensors such as a chest strap can be paired, and does not describe export to any third-party service — the companion app it names is Zepp and only Zepp. Amazfit's specification table also runs several Balance models in unlabeled adjacent columns, so figures that appear in only one column are not quoted here.

  6. [6]Manufacturer claim only2026
    Amazfit — FAQ, “How accurate is the GPS on the Amazfit Balance 2?”

    Read 2026-11-01. Amazfit's own answer to the accuracy question is the phrase “industry-leading GPS accuracy”, supported by “dual-band support from six satellite systems” and a promise of a “fast and reliable connection, even in challenging environments”. No methodology, no reference instrument, no measured figure and no named test appear anywhere on the page. This is the textbook case for the bottom two tiers of our evidence hierarchy: a claim about accuracy, published by the manufacturer, with nothing attached that anyone outside the company could check.

  7. [7]Manufacturer specification2026
    Garmin — Forerunner 165 Series owner’s manual, battery life information and specifications

    Manufacturer-stated, read 2026-10-11 from Garmin's own owner's manual. Battery: up to 11 days in smartwatch mode with activity tracking, phone notifications and wrist heart rate; up to 19 h in activity mode with GPS Only; up to 17 h with All Systems GPS; up to 25 h with UltraTrac; up to 7 h GPS Only with music playback and 6.5 h All Systems with music. The specification page states a rechargeable built-in lithium-ion battery, a 5 ATM swim water rating, up to 4 GB of media storage, an operating range of −20 to 60 °C and a charging range of 0 to 45 °C. NOTE the structure of the table: the highest-accuracy row Garmin publishes for this watch is All Systems — there is no all-systems multi-band row, because the hardware has no multi-band mode.

Read next