How to Measure Body Fat and Muscle Mass

The most accurate way to measure body fat and muscle mass depends on how much precision you need, what equipment you can access, and how much you’re willing to spend. Clinical gold standards like dual-energy X-ray absorptiometry (DEXA) and cross-sectional imaging with CT or MRI can pin down your body composition to within a few percentage points of your true values, while consumer tools like smart scales and skinfold calipers trade accuracy for convenience and cost. No single method is perfect for everyone, and each comes with trade-offs that matter more than most people realize.

DEXA Scans

DEXA is the method most often recommended by clinicians and researchers when they need a reliable snapshot of someone’s body composition. The scan takes about ten minutes, involves lying still on a padded table, and delivers separate estimates for fat mass, lean mass, and bone mineral content across your whole body and individual regions (arms, legs, trunk). The radiation dose is extremely low, comparable to a day or two of natural background exposure.

In validation studies comparing DEXA to chemical analysis of animal carcasses, the method proved accurate for soft-tissue composition, with measurement errors for percent fat around 3% and errors for lean body mass around 2.7 kg.¹ That level of precision is strong enough for tracking real changes over time, making DEXA a staple in sarcopenia research, obesity treatment programs, and sports science.

One important limitation: DEXA measures lean mass, not muscle mass directly. Lean tissue includes water, connective tissue, and organs alongside skeletal muscle. Researchers often use “appendicular lean mass,” the lean tissue in your arms and legs, as a proxy for muscle mass since those limbs are mostly muscle.² It’s a reasonable approximation, but it can be thrown off by fluid shifts, which is something we’ll get to shortly.

Bioelectrical Impedance Analysis

Bioelectrical impedance analysis, or BIA, works by sending a small electrical current through your body and measuring how quickly it travels. Fat tissue resists the current more than water-rich lean tissue, so the device uses that resistance (impedance) along with your height, weight, age, and sex to estimate body composition. BIA is the technology inside most consumer body-composition scales and many clinical body analyzers.

The accuracy gap between basic consumer BIA devices and more advanced systems is substantial. Simple single-frequency devices send current at one frequency and provide a rough estimate. Multi-frequency devices and bioelectrical impedance spectroscopy, which sweep across a range of frequencies, do a better job distinguishing intracellular from extracellular water. In elderly adults, multi-frequency BIA and spectroscopy were significantly better than single-frequency BIA at assessing skeletal muscle and diagnosing muscle weakness.³ If you’re choosing a BIA device for anything beyond casual curiosity, multi-frequency models are worth the extra cost.

BIA’s biggest practical weakness is sensitivity to hydration. Five hours of passive dehydration (just sitting in a warm room without drinking) was enough to shift body fat estimates by about a kilogram and alter phase angle readings.⁴ That means your reading on a Monday morning after coffee and a workout can look very different from a Tuesday morning reading after a long sleep with no water. If you use BIA for tracking, measure at the same time of day under the same conditions every time.

Skinfold Calipers

Skinfold measurements are the oldest portable body-fat method still in wide use. A trained tester pinches your skin at specific sites (commonly the chest, abdomen, and thigh for men; the triceps, hip, and thigh for women), measures the fold thickness with calipers, and plugs the numbers into prediction equations.

The accuracy depends heavily on who is doing the pinching and which equations they use. The widely used Jackson-Pollock equations were developed primarily on younger white adults, and their accuracy drops off in specific populations. In Black adults, one commonly used equation set underpredicted body fat by about 3 percentage points, while another overpredicted it by about 1.7 points, compared to a reference method using deuterium dilution. In white adults, overprediction was even larger with one equation set, reaching nearly 5 percentage points.⁵ The caliper brand also matters: different calipers apply different jaw pressures, and switching brands between measurements can introduce error on its own.

For people with obesity, traditional prediction equations tend to underestimate body fat. Modified equations using exponential models have been shown to cut that error roughly in half, bringing estimates within about 1% of DEXA values rather than the 2% gap seen with the original models for people with high skinfold sums.⁶ Still, skinfolds are best treated as a rough tracking tool rather than an absolute measurement. If the same trained person uses the same calipers on you every time, the trend over months is useful. The individual readings are not gospel.

Underwater Weighing and Air Displacement

Underwater weighing, or hydrostatic weighing, was long considered the reference standard for body density measurement. You exhale fully and submerge in a tank while a scale records your underwater weight. Because fat is less dense than lean tissue, your underwater weight relative to your dry weight reveals your body density, which prediction equations then convert to a body fat estimate.

The method is accurate but genuinely unpleasant for most people. You need to blow out as much air as possible while sitting underwater, and any residual lung volume introduces error. Air displacement plethysmography, marketed under the BOD POD brand name, replaces water with air. You sit inside a sealed chamber while the device measures how much air your body displaces. A validation study in subjects ranging from normal weight to severely obese compared this approach to underwater weighing and found the two methods produced comparable body density measurements.⁷ The BOD POD is more comfortable, faster (a few minutes versus the repeated dunking of underwater weighing), and available at many universities and sports medicine clinics.

Both methods measure body density, which is a single number. Converting density to body fat requires assumptions about the density of your lean tissue. The classic equations assume lean tissue has a fixed density based on three male cadavers from the mid-twentieth century. In practice, real people’s lean tissue density varies. A study of trained and sedentary men and women found that the mean density of fat-free mass was higher than the assumed standard, primarily because participants’ body water content was lower than the assumed value.⁸ This means a simple two-compartment model (fat versus everything else) can be off by a couple of percentage points. Adding a measurement of total body water, making it a three-compartment model, removed most of that error. Adding bone mineral data on top, making it a four-compartment model, offered only a tiny additional improvement of about 0.2 percentage points.

CT and MRI Imaging

If you need to know exactly how much muscle sits at a specific location in your body, or how much fat has infiltrated that muscle, cross-sectional imaging with computed tomography or magnetic resonance imaging is the most detailed option available. Both can distinguish visceral fat from subcutaneous fat, measure individual muscle cross-sectional areas, and detect fat marbling within muscle tissue.

CT and MRI produce highly consistent results when measuring the same structures. Skeletal muscle area measurements from the two modalities correlate strongly, with CT reading slightly higher on average by about 3.6 square centimeters.⁹ For fat compartments, they also agree closely, with MRI slightly underestimating visceral fat by about 3% relative to CT, and the two being nearly identical for subcutaneous fat.¹⁰

For most healthy people trying to track fitness progress, CT and MRI are overkill. They’re expensive, CT involves meaningful radiation, and a full body-composition scan requires specialized research protocols rather than a standard clinical order. These methods earn their place in clinical settings: monitoring cancer patients for muscle wasting, evaluating liver transplant candidates, or diagnosing sarcopenia in older adults when other methods give ambiguous results.

Ultrasound

Diagnostic ultrasound is emerging as a practical middle ground between the convenience of skinfold calipers and the precision of imaging methods. A trained operator places a probe against the skin and measures the thickness of subcutaneous fat and underlying muscle in real time. The equipment is portable, involves no radiation, and provides site-specific measurements that skinfolds and BIA cannot match.

For measuring subcutaneous fat and overall muscle thickness, standardized ultrasound protocols have been refined to improve reproducibility. The USVALID study protocol, for instance, measures fat and muscle at specific sites on the thigh and upper arm using a carefully defined procedure from skin surface to bone.¹¹ In sarcopenic obesity research, ultrasound measurements of the rectus femoris muscle in the thigh correlated meaningfully with grip strength, suggesting the technique captures functionally relevant muscle information, not just anatomy.¹²

Ultrasound also proves useful for distinguishing visceral from subcutaneous abdominal fat. Measurements of visceral fat depth by ultrasound correlate well with MRI reference measurements, particularly in older adults where the distinction matters most for metabolic risk.¹³ One significant advantage over BIA: ultrasound measurements of subcutaneous fat are essentially immune to hydration changes. When subjects were dehydrated by about 2% of body weight and then hyperhydrated by a similar amount, ultrasound fat-thickness readings barely budged, staying within 0.9 mm and less than half a percentage point of body fat across all conditions.¹⁴ That robustness makes ultrasound a strong choice when you need repeatable readings and cannot guarantee your hydration state will be identical each session.

Why Hydration Matters More Than You Think

The hydration sensitivity of BIA is well known, but it’s worth understanding that DEXA is not entirely immune either. When fluid balance changes, DEXA’s estimates of soft-tissue composition develop small but systematic errors. The magnitude depends on the type of fluid (its electrolyte content), how much fluid has shifted, and the person’s starting body composition.¹⁵ For a clinical scan where you’re well-hydrated and haven’t just finished a marathon, the errors are small enough to ignore. But for athletes weighing in after cutting weight, or patients with edema, the reading may not reflect true fat or lean tissue changes.

The practical takeaway: if you’re comparing two measurements taken weeks or months apart, try to be in roughly the same hydration state both times. For BIA that means same time of day, same relationship to meals and exercise, same approximate fluid intake. For DEXA the effect is smaller but still present.

Consumer Smart Scales and Their Limits

Bathroom smart scales that claim to measure body fat and muscle mass are everywhere, and the question most people have is whether the numbers mean anything. The honest answer is that the weight reading is reliable but the composition numbers are rough at best. A study comparing three commercially available smart scales to DEXA found that all three measured body weight within about half a kilogram, which is perfectly adequate. For fat mass, all three underestimated the true value, with median errors ranging from about 2 to 4.4 kg. For muscle mass, the picture was even less consistent: one scale underestimated by about 6.6 kg while the other two overestimated by around 4 to 4.5 kg.¹⁶

Those errors are large enough that a single reading is not clinically meaningful. A scale telling you that you’re 22% body fat when DEXA says 28% is not useful for making health decisions. Where smart scales might have value is in trend analysis: watching the direction of change over weeks and months rather than trusting any single number. Even there, researchers have noted that the small number of studies using connected smart scales makes it hard to validate the trend-tracking claim rigorously.¹⁷ If you own one, keep weighing yourself, but don’t make dietary or training decisions based on the body-fat percentage it displays.

Muscle Quality, Not Just Quantity

A growing area of body composition science goes beyond how much muscle you have to ask how healthy that muscle is. Myosteatosis, the infiltration of fat into and around muscle fibers, reduces muscle quality and is linked to weaker physical function and worse metabolic outcomes even when total muscle area looks adequate on a scan.¹⁸

CT scans can quantify this directly. Researchers divide total muscle area into “normal attenuation” muscle (healthy, dense, low-fat tissue) and “low attenuation” muscle (fattier, less dense). Low attenuation muscle area, measured at the mid-thigh level, has been consistently associated with reduced physical functioning and poorer metabolic prognosis.¹⁹ Automated tools can now generate color-coded muscle quality maps from a single CT slice, dividing muscle into healthy, low-quality, and intermuscular fat zones using density thresholds.²⁰

This matters because two people can have the same amount of muscle on a DEXA scan but very different functional capacity. If you’re an older adult and your doctor says your muscle mass looks fine but you still feel weak, muscle quality may be part of the story. Currently there’s no consumer-level way to assess myosteatosis; it requires CT or MRI. But the concept is worth knowing about, especially as it becomes a more routine part of clinical assessments in aging and oncology.

Tracking Changes Over Time

For many people, the real question is not “what is my body fat percentage right now?” but “is my body composition improving?” Tracking changes demands a method with high precision, meaning it gives you the same result when nothing has actually changed, so that when it does show a difference, you can trust it’s real.

CT imaging at the thigh level achieves a precision error of about 0.26%, meaning a change larger than about 0.7% in muscle cross-sectional area is likely real. At the abdomen, the precision error is a bit higher at 0.63%, with real changes needing to exceed about 1.8%.²¹ These are tight windows, which is why CT-based tracking is used in cancer treatment monitoring where detecting early muscle loss can influence clinical decisions.

Newer approaches like three-dimensional optical body scanning, which uses infrared cameras to create a surface model of your body, are being tested for longitudinal tracking. In a pilot study comparing 3D optical imaging to DEXA for monitoring growth and body composition changes in children, the optical method correctly identified 75% to 80% of the changes that DEXA flagged as significant.²² That’s promising but not yet in the same league as DEXA or CT for precision. These scanners estimate body composition from surface geometry using prediction equations, and validation studies show strong correlations to reference methods for whole-body fat mass and fat-free mass, with body-surface measurements like waist circumference matching closely to tape-measured values.²³

Distinguishing Visceral From Subcutaneous Fat

Total body fat percentage tells you how much fat you carry, but not where. Visceral fat, the fat packed around your organs inside the abdominal cavity, poses substantially greater metabolic risk than subcutaneous fat, the layer you can pinch under your skin. Separating the two requires methods that can see inside the abdomen.

MRI is the most accurate way to quantify visceral fat without radiation. CT is equally precise but involves ionizing radiation, making it less suitable for repeat measurements in healthy people. For settings where neither MRI nor CT is practical, a combination approach using BIA and ultrasound has been validated against MRI. In one study, this combined method estimated visceral fat mass with average errors of about 0.03 kg in women and essentially zero in men, though individual measurements could vary by over a kilogram in either direction.²⁴ That’s a practical workaround for epidemiological research or clinical screening when full imaging isn’t available.

Choosing a Method at Different Life Stages

Body composition measurement is not one-size-fits-all across the lifespan. In infants and young children, many standard methods simply don’t work well. Small body size makes DEXA positioning difficult, BIA reference equations are poorly validated, and skinfold measurements in babies require specialized training. A National Institutes of Health workshop reviewed the available techniques for children from birth through five years and identified significant gaps, noting that air displacement plethysmography (using the PEA POD, a smaller version of the BOD POD) and DEXA are among the best-validated options in this age range, but important accuracy questions remain.²⁵

In older adults, the priorities shift. The concern is often sarcopenia: the loss of muscle mass and strength that accelerates with aging and raises the risk of falls, fractures, and disability. Diagnosing sarcopenia involves measuring muscle mass (typically with DEXA or BIA), muscle strength (usually grip strength), and physical performance (walking speed or chair-stand tests).²⁶ The choice of how you adjust the muscle mass number for body size also changes the diagnosis. Dividing appendicular lean mass by height squared, by total body weight, or by BMI each yields a different prevalence of sarcopenia and different clinical implications.²⁷ This means two doctors using different cutoffs may disagree on whether the same person has sarcopenia. For screening purposes, BIA-derived fat-free mass index cutoffs of roughly 18 kg/m² for men and 15 kg/m² for women have been proposed as practical thresholds for flagging low muscle mass.²⁸

3D Optical Body Scanning

Three-dimensional optical scanning is one of the newest approaches to reach consumers. You stand in front of one or more infrared depth cameras, the device captures your body shape in seconds, and software extracts circumferences, volumes, and surface area. From those shape measurements, prediction equations estimate fat and lean mass regionally and for the whole body. Several commercial scanners are now available in gyms, wellness clinics, and even as smartphone-based apps.

Validation studies against four-compartment reference models have been conducted for multiple commercially available scanners.²⁹ The technology produces highly reproducible body-shape measurements, with surface-area and volume correlations above 0.97 when compared to reference methods. Predicted fat mass and fat-free mass at the whole-body level show strong agreement, with root mean square errors of about 2.2 to 2.4 kg.²³ The appeal is speed, zero radiation, and the ability to visualize where your body is changing shape. The limitation is that composition estimates are inferred from surface geometry, not measured directly inside the body. If two people have the same external shape but different internal fat distribution, the scanner may not tell them apart.

The field of body composition measurement has come a long way from the early twentieth century, when knowledge of human composition relied mainly on cadaver studies and chemical analyses of isolated organs.³⁰ Today’s tools can quantify body components at the molecular, cellular, and tissue levels in a living person. For the average person deciding how to measure their own body fat and muscle, the landscape simplifies: use DEXA or a BOD POD if you want accuracy, use a well-controlled BIA or ultrasound protocol if you want convenience with reasonable precision, and treat consumer scale readings as entertainment rather than clinical data. Whatever method you choose, consistency in how and when you measure will always matter more than which device you pick.