DEXA vs. InBody: which is more accurate for body fat?

DEXA (also written DXA) and InBody are two different ways clinics, gyms and wellness providers estimate body composition: the fat mass, lean mass and, for DEXA, bone mineral. This comparison of a DEXA scan vs. bioelectrical impedance draws on research that measured each method against an independent reference method, and on studies that used an InBody device specifically.

Quick comparison

Each figure keeps the study population and comparator it was reported with; see the DEXA in depth and InBody / bioimpedance in depth sections for those details. Accessed 2026-09-27.

These are estimates, and a licensed clinician interprets the results [9].

DEXA scanners are regulated by FDA as medical devices (Class II). The Body Composition Software Option for GE Lunar DEXA Bone Densitometers (GE Medical Systems Lunar, K071570) was cleared through FDA's 510(k) process for prescription use. Its cleared indications say the software 'measures the regional and whole body bone mineral density (BMD), lean and fat tissue mass and calculates derivative values' such as body-fat percentage, and that it 'does not diagnose disease, or recommend treatment regimens, or quantify treatment effectiveness. Only the health care professional can make these judgments.'

DEXA in depth

DEXA uses two X-ray energy levels to separate bone mineral from soft tissue, then soft tissue into fat and lean mass, across the whole body or by region [2]. These values are estimates, and the software labeling quoted above says it does not diagnose disease [9].

DEXA scanners are FDA-regulated medical devices (Class II). Manufacturers market them through FDA's 510(k) clearance process.

Clearance is not the same as approval: FDA's own rule states that a 510(k) determination "does not in any way denote official approval of the device" [10]. The clearance quoted above covers one software option for one manufacturer's scanners, not every DEXA scanner or software version on the market.

A DEXA bone-density exam uses approximately 0.001 mSv on older pencil-beam scanners and about 0.009 mSv (hip) to 0.013 mSv (spine) on fan-beam Hologic scanners, according to a 2010 review [7]; a whole-body body-composition scan uses under 10 microsieverts (0.01 mSv), which researchers compare to about one day of natural background radiation at sea level [2]. Tell the facility if you are or may be pregnant before a DEXA scan [11].

If you have a medical emergency, call 911. Problems with medical devices, including imaging equipment, can be reported to FDA through MedWatch.

Regulatory status checked September 27, 2026. Sources: www.govinfo.gov/…/CFR-2025-title21-vol8-sec892-1170.xml; www.accessdata.fda.gov/…/K252718.pdf; www.accessdata.fda.gov/…/K071570.pdf; www.govinfo.gov/…/CFR-2025-title21-vol8-sec807-97.xml; www.fda.gov/…/it-really-fda-approved

InBody / bioimpedance in depth

InBody is a brand of bioelectrical impedance analysis (BIA) device. BIA estimates body composition by passing a small electrical current through the body, measuring resistance, and converting that reading to estimated body water, fat-free mass and fat mass using the device's own equations [3]. Most of the research behind this comparison studied "bioimpedance technologies" generally, without naming a device brand; this page names InBody only where the source it cites used that brand.

In the same study of 170 healthy adults, the bioimpedance devices tested had 95% limits of agreement of 6.2 percentage points or less, wider than DEXA's, although their average error was small; the authors describe them as performing well but with "larger individual-level errors" [1]. A separate systematic review of 12 studies comparing BIA with a 4-compartment reference model in healthy adults found mean bias for percent body fat ranging from −3.5% to +4.4%, with limits of agreement typically spanning 15 to 20 percentage points; the authors concluded BIA estimates were "not equivalent" to the reference model at the individual level, and noted that most BIA equations are proprietary [4]. Because the 170-adult study and this review used different reference models (5-component vs. 4-compartment) and may report limits of agreement differently, their figures should not be compared directly.

One study using an InBody 770 device specifically, in 82 young, normal-weight adults, found a 4.2-percentage-point total error for percent body fat against a 4-compartment model; the authors wrote that stand-up multi-frequency BIA "may overestimate %fat and FM, and underestimate FFM" compared with that model [12]. We have not yet checked InBody's FDA status, so this page does not state it. We did not review safety cautions for bioimpedance devices; ask the facility whether the test is suitable for you.

Repeatability: comparing results over time

In one study of 30 weight-stable adults on a GE Lunar iDXA scanner, same-day repeat scans varied by less than 2% for every measure except visceral fat [5]. Results from different scanners are not directly comparable, even between models from the same manufacturer: in one 2026 study of 101 adults scanned the same day on three scanners, converting body-fat percentage between devices left biases from about 1 percentage point (GE iDXA to Hologic Discovery) to about 4.2 points (Hologic Discovery to Hologic Horizon) [13].

The International Society for Clinical Densitometry states that no phantom has been identified that removes systematic body-composition differences between DEXA manufacturers [14]. For body composition, ISCD sets minimum precision standards for each technologist of 3% for total fat mass, 2% for total lean mass and 2% for percent body fat [14]. In one GE iDXA study, a change in body-fat percentage smaller than about 4.4 percentage points across 3 months could be measurement noise (95% least significant change, 51 weight-stable adults) [5]. The InBody-specific repeat-measurement figure given on this page comes from the no-fluid arm of a small pilot: body-fat readings fluctuated by up to 0.7 percentage points over 2 hours (InBody 770, 18 adults) [6]. We did not search for repeatability studies on InBody devices for this page, so no figure for repeat scans over weeks or months is given.

Hydration and fluid sensitivity

Recent fluid intake shifted bioimpedance readings in the studies reviewed here. In one study of 76 healthy young adults using segmental BIA, percent body fat rose by about 1.1 to 1.2 percentage points 20 minutes after drinking 591 mL of water or a carbohydrate-electrolyte drink [8]. DEXA lean-mass readings can also be affected by hydration status [3]. In one randomized trial of 46 adults, drinking 1000 mL of water raised DXA lean soft tissue estimates compared with 0, 250 and 500 mL, mostly in the trunk [15]. This page does not compare the size of the effect between DEXA and bioimpedance.

A small pilot study using an InBody 770 device (18 adults, crossover design) tested several drinks. Fat mass rose significantly after glucose and after Ringer's solution, but not after saline or no fluid. The effect of time depended on the drink (a significant drink-by-time interaction), while drink type alone was not significant. The authors called the clinical relevance uncertain because the mean differences were below 1%. Even with no fluid, body-fat readings fluctuated by up to 0.7% over 2 hours [6]. The authors of one bioimpedance study recommend following pre-test fluid-restriction guidance [8]. Ask the facility whether it has other instructions about food and fluid before either test.

Which might make sense for different goals

Choosing between a DEXA scan and an InBody (or similar bioimpedance) test depends on what it's for, not on which one is universally more accurate. Each method has strengths this page's comparisons support, and gaps this page did not research for either method: cost, availability and wait time.

• Tracking body-fat change over time: repeatability on the same device matters most. On one GE iDXA scanner, same-day repeat scans varied by under 2% for most measures [5]; this page gives only a 2-hour figure for InBody devices (18 adults) [6] and did not search for longer-term repeatability studies. For either method, compare results from the same device under similar conditions [5][13].
• Wanting a bone-density result at the same visit: a DEXA scan can include a bone-density result; this page did not look for bone estimates from bioimpedance devices. Bone density is a separate, established use of DEXA with its own screening guidance.
• Avoiding X-ray exposure entirely, even at DEXA's very low dose: bioimpedance devices such as InBody use no ionizing radiation.
• Comparing results across visits with different recent fluid intake: bioimpedance readings shifted after drinking in the studies reviewed here [8][6]. DEXA lean-mass estimates have also shifted after water intake in a randomized trial [15].
• Cost and where each is offered: this page did not research pricing or availability for either method.

A licensed clinician or the specific facility can advise which test fits your situation.

Limitations of the evidence

The comparisons on this page come from a small number of studies, mostly in healthy adults, and results may not apply to other groups. The one study reviewed here that tested DEXA and bioimpedance against the same reference method studied 170 healthy adults; the InBody 770 accuracy study studied 82 young, normal-weight adults, and the InBody 770 fluid-intake pilot studied 18 adults [1][12][6].

This page does not cover whether tracking body fat or lean mass with DEXA or InBody changes health outcomes in healthy adults, and we did not find a trial that tested it (Europe PMC, PubMed and ClinicalTrials.gov, searched 2026-10-01). A 2020 single-center, open-label randomized trial of bioimpedance-guided fluid management in 240 people on peritoneal dialysis, funded by public grants in China, found no significant difference in 1-year survival (11 deaths in total), which is a different question [16]. Most bioimpedance equations are proprietary and vary by device and manufacturer, and DEXA results can also differ by scanner model, so a change in reading over time is more meaningful when it comes from the same device [4][13].

Questions to ask a provider

Before scheduling a DEXA scan or an InBody test, ask the provider which method they use, which device or scanner model, and whether they can compare your results to a previous visit on the same machine. These details affect how much weight to put on any change between two scans.

• Which method do you use, and which device or scanner model?
• Will my next scan be on the same machine, at the same facility?
• Do you standardize conditions (recent food and fluid intake, clothing, time of day) before scanning?
• What does the price include, and does your facility require a clinician's order?
• Can I see my raw numbers, not just a category or score?

Nearby providers

Some facilities require a clinician's order; requirements vary by state and facility. Ask the facility what it requires before you schedule.

Find a Provider

Related diagnostics and treatments

• Diagnostic hub: DEXA Scan
• Comparison: DEXA Body Composition vs. Bone Density Scan