How to Measure Lumbar Lordosis: Methods and Normal Range

Lumbar lordosis is measured most accurately with a standing lateral X-ray and a technique called the Cobb angle, which captures the inward curve of the lower spine between the top of the first lumbar vertebra and the top of the sacrum. For the average adult, that angle falls somewhere in the range of roughly 30° to 60°, though “normal” depends heavily on your sex, age, pelvic anatomy, and even ethnicity. Several non-radiographic tools exist for clinical and research settings, each with trade-offs in accuracy and convenience that are worth understanding before you trust any single number.

The Cobb Angle on X-Ray

The standard approach to measuring lumbar lordosis uses a standing lateral radiograph. The clinician draws a line along the top endplate of the L1 vertebra and another along the top endplate of S1, then measures the angle formed by perpendiculars dropped from those two lines. This L1–S1 Cobb angle is the most widely reported measurement in the spine literature and consistently shows the highest diagnostic accuracy when classifying someone as having too much curve (hyperlordosis) or too little (hypolordosis).¹

A narrower version, the L1–L5 Cobb angle, stops one level short by drawing the lower line at L5 instead of S1. It leaves out the lumbosacral junction, which is significant because the L5–S1 segment contributes a large share of the total curve. In studies comparing several radiographic techniques, both the L1–S1 and L1–L5 Cobb methods showed high inter- and intra-observer reliability, with mean absolute differences between observers in the range of about 1° to 4°.² The L1–L5 method tends to perform especially well for inter-observer agreement across different spine conditions, making it a practical fallback when S1 is hard to identify on the film.³

Other Radiographic Techniques

Beyond the Cobb angle, researchers have developed alternatives like the posterior tangent method, the centroid method, and the TRALL (tangent–radius–angle of lumbar lordosis) method. Each uses different anatomical landmarks or geometric constructions to describe the curve. These alternatives can capture additional information, such as the shape of the curve rather than just the total angle. The posterior tangent method, for instance, draws tangent lines along the back wall of each vertebral body, producing both segmental and global measurements of lordosis.

In practice, the choice of method matters less than using the same method consistently. All the main techniques produce reliable measurements within a single observer, but the absolute numbers differ between methods because they define the curve geometry differently. A lordosis measured by the posterior tangent method will not match a Cobb angle from the same X-ray. For comparing a patient’s measurements over time, or comparing one patient with published norms, you need to know which method was used. The Cobb L1–S1 angle remains the benchmark that most published normal values reference.⁴

Measuring Without Radiation

Not every assessment needs an X-ray. In physical therapy clinics and research labs, two non-invasive tools are commonly used: the flexible ruler (flexicurve) and the digital inclinometer.

A flexicurve is a bendable metal-and-rubber strip that a therapist molds against your spine while you stand. The therapist traces the shape onto paper and calculates the lordosis from the depth and length of the curve. Studies have found that a single therapist can get repeatable measurements this way, with intraclass correlation coefficients above 0.80 in many cases. The catch is that different therapists measuring the same person often disagree substantially, with inter-tester reliability values dropping as low as 0.41 to 0.50.⁵ So if your physical therapist is tracking your lordosis over several visits, the readings are useful as long as the same person takes them each time. Comparing numbers between two different clinicians is risky.

Digital inclinometers work differently: you place the device at the top and bottom of the lumbar spine and read out the tilt angle at each level. The difference gives the lordosis. Optimizing how many readings you take per session (triplicate measurements, for example) and having consistent rater protocols can improve reliability, as demonstrated in studies applying measurement theory to postmenopausal women with spinal osteoporosis.⁶ The practical takeaway is the same: one examiner, standardized protocol, multiple readings averaged together.

Surface Topography and Rasterstereography

A more recent option is rasterstereography, which projects a grid of light onto your back and uses cameras to create a 3D surface map of your spine. The technology is entirely radiation-free and captures the measurement in seconds. A meta-analysis found that rasterstereography has strong reliability for assessing lumbar lordosis (pooled correlation above 0.91) and reasonable validity when compared against radiographic imaging (correlation above 0.70).⁷

That said, the correlation with X-ray is not perfect, and when researchers have directly compared rasterstereographic scans with clinical orthopedic examinations, statistically significant differences in lumbar lordosis readings emerged.⁸ The technology works well for screening, monitoring change over time, and reducing cumulative radiation exposure in patients who need frequent follow-up. It is not yet a direct replacement for radiographs when precise surgical planning is on the table.

AI-Assisted Measurement

Measuring Cobb angles by hand on an X-ray takes time and introduces human variability. Deep learning models have been trained to identify vertebral landmarks automatically and calculate lordosis angles from plain lateral radiographs. In one recent study, the agreement between an AI model and two orthopedic surgeons was excellent, with an overall intraclass correlation of 0.962 and Pearson correlations between the AI and each surgeon above 0.85.⁹ Another model trained to measure lumbosacral parameters showed accuracy on par with attending-level radiologists while being faster and more reproducible.¹⁰ These tools are not widely deployed in community clinics yet, but they are moving quickly from research prototypes toward routine use, and they address one of the biggest practical problems in spinal measurement: the variability between different human readers.

What Counts as a Normal Lordosis Angle

Published normal values vary because different studies use different populations, age ranges, and measurement methods. One widely cited study of 149 healthy adults found a mean lumbar lordotic angle of about 33° with a standard deviation of 12°, suggesting a “normal” range of roughly 20° to 45° when using one standard deviation as the boundary.¹¹ A much larger study of 2,500 healthy individuals in Iran, using surface measurement rather than X-ray, found a mean of about 42° with a total range spanning from around 10° to 75°.¹² An EOS imaging study reported a mean lordosis angle of about 32° with a narrower spread.¹³

The spread across these studies reflects genuine biological variation, but it also reflects the measurement tools. Surface-based methods and radiographic methods will not produce identical numbers for the same spine. If your report says your lordosis is 35° and a textbook says normal is 45°, the discrepancy could simply be a difference in technique rather than a sign that something is wrong. Always compare your number to norms derived from the same method.

Why Sex, Age, and Body Composition Shift the Numbers

Women tend to have greater lumbar lordosis than men. One anatomical study found the lumbar angle was about 7° larger in women when standing, a difference that disappeared when lying down, suggesting both vertebral shape and postural factors are at play.¹⁴ The female lumbar curve also tends to peak lower in the spine, and the upper portion of the curve tilts more posteriorly, which contributes to the visual impression of deeper lordosis.¹⁵ These differences are thought to be an evolutionary adaptation that helped redistribute load during pregnancy.

Age matters too. Lordosis in children increases steadily from about 30° in toddlers to around 44° by the late teens.¹⁶ Infants start with almost no lordosis at all; the curve develops in tandem with learning to walk and stand upright.¹⁷ In older adults, degenerative changes in the discs and facet joints tend to flatten the curve again.¹⁸

Body fat distribution plays a role as well. In a study of a large Caucasian population, lordosis angles increased with higher body adiposity, though the correlation was weak.¹⁹ The large Iranian surface-measurement study confirmed a significant three-way interaction among sex, age, and body mass index on lordosis values.¹² And ethnicity adds another layer: a comparison of spinal alignment across Mexican, Asian, and Caucasian groups found mean lumbar lordosis values of about 60°, 52°, and 61° respectively, with the difference between the Mexican and Asian populations being statistically significant.²⁰

How Your Position Changes the Reading

Lordosis is not a fixed number. The same spine produces different measurements depending on whether you are standing, sitting, or lying down. Lordosis is generally greatest when standing and decreases as you move to sitting or supine positions. In one study of middle-aged patients, lumbar lordosis dropped significantly from standing to sitting, with the L5–S1 segment contributing the largest share of that change.²¹ Another study measured lordosis in both young and older groups and found standing values around 53° in both, with progressive flattening when sitting.²²

An interesting finding from MRI-based shape modeling is that although posture changes the total amount of curvature, each person’s spine retains a characteristic shape signature across positions. The relative distribution of the curve shifts, but not randomly: the spine is straightest when standing and shows the most evenly distributed curvature when sitting.²³ For clinical comparisons, this means standing lateral films are the standard for good reason. Any other position introduces systematic differences that make it hard to compare with published norms.

Even arm position during a standing X-ray can influence the reading. Having the arms abducted or raised to different positions shifts the balance of the trunk slightly, nudging kyphosis and lordosis values up or down. These changes are generally small and not statistically significant in most studies, but standardized arm positioning is still part of good radiographic protocol for spinal alignment.²⁴

Why Pelvic Incidence Matters More Than You’d Expect

Your lumbar lordosis does not exist in isolation. It is mechanically linked to the shape and tilt of your pelvis, and one of the most important relationships in spinal alignment is between lumbar lordosis and a measurement called pelvic incidence. Pelvic incidence is a fixed anatomical parameter that reflects the geometry of your pelvis; unlike lordosis, it does not change with posture. It essentially sets a “budget” for how much lordosis your lumbar spine needs to maintain balanced upright alignment. A well-aligned spine keeps lumbar lordosis within about 10° to 11° of the pelvic incidence.²⁵

What this means in practice is that two people can have very different lordosis angles and both be perfectly aligned. A person with a high pelvic incidence naturally needs and has a deeper lordosis; a person with a low pelvic incidence has a flatter curve. Judging either against a single “normal range” misses the point. The lordosis-to-pelvic-incidence match is what determines whether the spine is balanced. Research has shown that the upper portion of the lumbar curve (proximal lordosis) is the segment most responsive to pelvic incidence, while the lower portion remains relatively constant across individuals.²⁶

Lordosis and Low Back Pain

The relationship between how much curve you have and whether your back hurts is more complicated than many people assume. A systematic review and meta-analysis found that people with low back pain tended to have slightly less lordosis than pain-free controls, and the difference was most pronounced in patients whose pain was related to disc herniation or degeneration.²⁷ That finding suggests reduced lordosis is associated with disc problems, but the effect size in age-matched studies was modest.

Confusingly, too much lordosis can also be a problem. A study of prolonged standing found that people who developed pain during the protocol had about 4° more lordosis on average than those who stayed comfortable.²⁸ So both too much and too little curve can be associated with pain, depending on the context. The honest summary is that lordosis is one factor among many, and an “abnormal” number on a report does not by itself predict who will hurt and who will not.

Flat Back Syndrome and Degenerative Loss of Lordosis

When lordosis is markedly reduced, it can lead to a condition called flat back syndrome. People with this condition stand with their trunk pitched forward and often develop chronic back pain, fatigue in the spinal muscles, and difficulty maintaining an upright posture. The forward shift puts compressive forces mainly on the front of the vertebral bodies and discs rather than distributing load across the whole motion segment.²⁹ Over time this becomes a self-reinforcing cycle: the loss of lordosis causes increased disc loading, which causes further disc-height loss, which flattens the curve even more.¹⁸

Targeted exercise programs that focus on lumbar extension and core muscle activation have been studied for flat back syndrome. The goal is to strengthen the deep spinal muscles and encourage the spine to maintain whatever lordosis is structurally available.³⁰ When exercise alone is insufficient and the deformity is severe, surgical correction becomes a consideration.

Surgical Planning and Target Angles

Spine surgeons do not aim for a generic lordosis angle when correcting adult spinal deformity. The target is personalized to the patient’s pelvic incidence. One frequently referenced formula derives the target lumbar lordosis as roughly 0.45 times the pelvic incidence plus about 32°.³¹ Getting this wrong has consequences: overcorrecting lordosis by roughly 15° or more beyond the ideal target has been linked to increased mechanical complications after surgery, including hardware failure.³² Undercorrecting leaves the patient still pitched forward. This is why the lordosis-pelvic-incidence relationship described earlier is not just an academic curiosity; it directly determines surgical outcomes.

An Evolutionary Quirk Worth Knowing

Lumbar lordosis is one of the defining features of human anatomy. Other primates have only small lordosis angles, while humans display curves ranging from about 30° to 80°.³³ The lordosis positions the trunk’s center of mass directly over the hip joints, making upright walking energetically efficient.³⁴ Fossil evidence and modern comparative anatomy suggest that the evolution of a deeper, more flexible lumbar curve was a critical step in the transition to habitual bipedalism, and that the wide variability we see in living humans has existed throughout hominin history.³⁵ Interestingly, MRI studies have detected structural lordotic wedging in the L4 and L5 vertebrae of fetuses, meaning your spine begins preparing for upright posture before you are even born.¹⁷