General Movement Assessment: What Parents Should Know

General Movement Assessment, or GMA, is a method clinicians use to evaluate a young baby’s spontaneous movements as an early indicator of brain health, most commonly to screen for cerebral palsy. The assessment involves recording a short video of your baby moving freely, then having a trained observer analyze the quality and variety of those movements. It is non-invasive, requires no needles or sedation, and can be done in a clinic or even at home with a smartphone. For parents of infants who spent time in the neonatal intensive care unit or had a complicated birth, GMA is one of the earliest and most reliable tools available to determine whether your child’s nervous system is developing on track.

What Clinicians Are Actually Looking At

The assessment focuses on spontaneous movements your baby makes on their own, not reflexes triggered by a clinician tapping or prodding. A healthy infant’s spontaneous movements have a specific quality that trained observers learn to recognize: they involve the whole body, vary in speed and intensity, include subtle rotations, and flow smoothly from one movement pattern to another. Researchers describe these as having a rich and complex repertoire with variable sequences of arm, leg, neck, and trunk movements that come and go gradually.¹ Think of it like watching water flow in a stream: there is an organic quality to it, with constant small changes in direction and speed.

When something is off with the developing brain, these movement patterns change in recognizable ways. Movements might become stiff and happen simultaneously across the body rather than flowing in sequence. Or they might look monotonous, lacking the variation and complexity you see in a healthy infant. Trained assessors are not counting movements or measuring angles; they are reading the overall quality of the movement like you might read someone’s body language. This visual “Gestalt” approach, where the whole pattern matters more than any single detail, is what makes GMA distinctive.

The Two Assessment Windows

GMA covers two distinct periods in early infancy, and each period has its own type of normal movement pattern that clinicians look for.

The first window runs from the preterm period through roughly the first two months after your baby’s due date. During this phase, clinicians look at what are called “writhing movements,” the whole-body spontaneous movements described above. The most concerning pattern at this stage is called “cramped-synchronized,” where the limbs and trunk stiffen and move rigidly together rather than in the flowing, variable pattern of healthy movement.

The second window, from roughly 3 to 5 months after the due date, is often considered even more important. This is when healthy babies develop what are known as “fidgety movements,” small, elegant, circular movements of the neck, trunk, and limbs that happen continuously in the background while the baby is awake. In healthy full-term infants, these movements start out sporadic around 4 weeks post-term, become more continuous and refined between 11 and 16 weeks (with rotational movements appearing in the wrists and ankles), and then gradually fade away by about 18 to 20 weeks.² The absence of fidgety movements during this window is one of the strongest early predictors that a baby may develop cerebral palsy.

Why the Brain Connection Matters

What makes these movement patterns so revealing is that they are generated by the developing brain itself. Fidgety movements are thought to reflect a specific developmental transition in the brain, where cortical network activity shifts from widespread patterns to more fragmented, refined activity.³ When those movements look abnormal or are absent, it signals that extensive cortical and subcortical brain networks may not be developing as expected. This is why clinicians treat movement quality as a window into what is happening inside the brain, even before a baby is old enough for more complex neurological testing.

This also explains why GMA works so early. Traditional neurological exams rely on milestones like sitting, crawling, or walking, but those skills develop over months. Spontaneous movements are present from birth, and their quality reflects brain integrity from the very start. A baby whose movements are subtly abnormal at 3 months may look perfectly fine to a parent’s eye, but a trained assessor can pick up patterns that suggest the brain’s wiring is not quite right.

How Accurate Is It

This is usually the first thing parents want to know, and the honest answer is that GMA’s accuracy depends on context: what population it’s used in, which movement period is being assessed, and whether it’s combined with other tools.

Used on its own during the fidgety movement period, GMA is strong at ruling out cerebral palsy. In one study of high-risk infants, absent or sporadic fidgety movements had about 76% sensitivity (meaning it caught about three-quarters of the babies who went on to develop cerebral palsy) and about 82% specificity, with a negative predictive value near 97%, meaning that if your baby’s fidgety movements looked normal, there was a very high chance they did not have cerebral palsy.⁴ In term-born infants who received therapeutic cooling after birth complications, the absence of normal fidgety movements at three months was associated with cerebral palsy with about 89% sensitivity and 89% specificity.⁵

Those numbers improve when GMA is combined with brain imaging. In one study, using GMA together with brain MRI reached about 87% sensitivity, over 98% specificity, and roughly 98% overall accuracy for predicting cerebral palsy after age two.⁶ Another study of preterm and low-birth-weight infants found that combining MRI findings with GMA results reached over 92% sensitivity and nearly 98% specificity.⁷ And in a study of high-risk term neonates, neither abnormal MRI findings alone nor absent fidgety movements alone had great sensitivity on their own (both around 29%), but the combination was extremely good at ruling out moderate-to-severe cerebral palsy: no child with both normal MRI and normal fidgety movements developed the moderate or severe form of the condition.⁸

For parents, the practical takeaway is that a normal GMA result is very reassuring, especially when paired with normal imaging. An abnormal result, on the other hand, does not guarantee a diagnosis. The positive predictive value, the chance that an abnormal result truly means cerebral palsy, can be as low as a third in some clinical follow-up settings. That means many babies flagged by GMA will turn out to develop normally. Clinicians know this and use an abnormal GMA as a reason to monitor more closely and start early supports, not as a definitive diagnosis.

Combining GMA With Other Clinical Tools

GMA rarely operates in isolation. Clinicians often pair it with the Hammersmith Infant Neurological Examination (HINE), a structured exam that scores a baby’s posture, movements, tone, and reflexes at specific ages. In infants treated with therapeutic cooling after birth, HINE scores at 3, 6, and 9 months were associated with cerebral palsy, with sensitivity ranging from about 82% to 90% and specificity reaching as high as 100%.⁵ In another study, combining GMA and HINE reached 100% specificity for cerebral palsy prediction while maintaining strong sensitivity.⁹

The pattern across studies is consistent: no single tool is perfect, but layering GMA with HINE and brain MRI gives clinicians a much clearer picture than any one assessment alone. If your baby’s medical team mentions multiple types of assessment, that is not a sign that something is wrong. It is how the best follow-up programs work, using each tool to fill in the gaps the others leave.

What Happens After an Abnormal Result

This is where early identification pays off. The entire point of screening early is to begin interventions during the first months of life, when the brain is most adaptable. Evidence suggests that early intervention can improve neurological outcomes. In one study, premature infants who showed cramped-synchronized movement patterns at term age and received early intervention went on to develop normal fidgety movements, suggesting improvement in their neurological status.¹⁰

A systematic review of early intervention and spontaneous movements found that early intervention could influence movement quality, particularly during the fidgety period, though the review also noted that the evidence base has limitations: studies used different methods, and the overall certainty of the evidence was low.¹¹ This is an area where the science is still catching up to clinical practice. Most neonatal follow-up programs do not wait for perfect evidence before offering therapy, because the window for early brain plasticity is narrow and the interventions themselves, which typically involve physical therapy, sensory enrichment, and parent coaching, carry little risk.

For parents, an abnormal GMA result typically means your baby will be enrolled in closer developmental monitoring and offered early therapy services. The specific interventions vary by setting but often focus on helping you, the parent, learn how to position, handle, and play with your baby in ways that support motor development. You are not just a bystander in this process; parent-delivered interventions are central to most early therapy programs.

Recording the Assessment at Home

One of the most parent-friendly developments in GMA is that the video can now be recorded at home using a smartphone. Several apps and programs have been developed specifically for this purpose. In one multi-site study, parents were given a short instructional video and then asked to record their infant’s spontaneous movements at home. Nearly all of the returned videos, 96%, met the technical standards needed for a clinician to perform the assessment remotely.¹² Other smartphone-based programs, such as the Baby Moves app, have shown similar success in getting parents to capture usable video at home.¹³

Remote GMA has even been piloted in lower-income settings. A feasibility study in Nepal found that engaging parents in cerebral palsy screening through a smartphone application was possible in an urban area and was generally welcomed by the families who participated.¹⁴ This matters because many families live far from specialized neonatal follow-up clinics. The ability to film your baby at home and send the video for expert review removes a significant barrier to access.

If you are asked to record a video at home, the basics are straightforward: your baby should be awake, calm, lying on their back in a diaper or light clothing so limbs are visible, and not crying or feeding. The camera should be positioned above, capturing the whole body, and the recording typically needs to last a few minutes. The app or your clinician will provide specific instructions, but the key point is that this is something parents can do well with minimal training.

How Parents Experience the Process

Parents understandably worry about any kind of developmental screening. Will the result make me more anxious? What if I get bad news? Research into parents’ experiences has been surprisingly positive. In a qualitative study, parents described the early risk assessment process using a mobile application as rewarding rather than frightening. Rather than increasing worry, having a concrete way to participate in their child’s assessment gave parents a sense of security and control. Parents felt that knowing about any risk was better than not knowing, and the ability to choose which video clips best represented their child’s typical movements made them feel like active participants in the process rather than passive recipients of clinical judgment.¹⁵

This is worth emphasizing because many parents of NICU graduates already carry significant anxiety. Having a clear, understandable screening process with good information about what the results mean can actually reduce that anxiety rather than adding to it. If your baby’s team recommends GMA, it is not because they think something is wrong. It is standard practice for babies who had complicated beginnings, and the overwhelmingly likely outcome is reassurance.

Why Rater Training and Experience Matter

GMA depends on the assessor’s trained eye, which means the quality of the assessment is only as good as the person reading the video. Assessors go through formal training in what is known as the Prechtl method, named after Heinz Prechtl, who developed the approach. Training occurs in two levels: basic and advanced. Research has shown that completing advanced training significantly improves both the consistency between different raters and the accuracy of cerebral palsy prediction.¹⁶

Experience after training also makes a difference. One study found that agreement between experienced GMA-certified raters was high when distinguishing normal fidgety movements from other movement categories, but less experienced raters showed lower consistency, particularly when trying to distinguish subcategories of normal movements.¹⁷ A structured checklist tool has been developed to support standardization, and studies using it have shown good to excellent agreement between scorers for distinguishing normal from abnormal movements.¹

For parents, the practical implication is to ask whether the person reading your baby’s video has completed certified GMA training and ideally has substantial experience. In well-established neonatal follow-up programs, this is standard. In settings where GMA is newer, it is a reasonable question to ask.

Beyond Cerebral Palsy

While cerebral palsy prediction is the primary use of GMA, researchers have explored whether the same movement patterns carry information about other neurodevelopmental conditions. A review of early motor abnormalities in infants later diagnosed with autism spectrum disorder found an exceedingly high rate of abnormal general movements in those infants, though the authors cautioned that more research is needed before GMA can be considered a standalone screening tool for autism.¹⁸ The same review noted associations with Rett syndrome.

This is an area where parents should set expectations carefully. GMA was developed for and validated against cerebral palsy outcomes. Its potential relevance to autism or other conditions is an active area of research, not an established clinical application. If your clinician mentions GMA, the conversation is almost certainly about motor development and cerebral palsy risk, not about autism screening.

Artificial Intelligence and the Future of Screening

One of the biggest limitations of GMA is that it requires trained human assessors, and there are not nearly enough of them worldwide to screen every at-risk infant. Researchers have been working for years to automate the assessment using artificial intelligence. A systematic review of AI approaches to GMA found a growing field of research attempting to transfer the assessment from human visual perception to computer-based analysis, with deep learning approaches showing particular promise.¹⁹

More recently, an international consortium has been established specifically to develop a roadmap for implementing an AI-enabled GMA system for universal cerebral palsy screening worldwide.²⁰ The vision is ambitious: a system where parents record a video on their phone, the video is analyzed by AI, and results are relayed to their clinical team without needing a certified assessor to watch every clip. This is not yet available in routine clinical practice, but it is the direction the field is heading, and for families in regions without access to trained assessors, automated GMA could be transformative.

Access in Lower-Resource Settings

GMA’s appeal in low- and middle-income countries is straightforward: it requires only a video camera and a trained eye, no expensive equipment. A review of GMA use outside Europe found that studies in countries including China, India, Iran, South Africa, and Brazil have demonstrated the assessment’s value, and the authors described GMA as ideal for assessing young nervous system integrity in low-resource settings due to its high accuracy and minimal cost.²¹

Implementation work is bearing fruit. In one program in a low-resource setting, GMA was successfully embedded into a routine high-risk newborn follow-up clinic. Three-quarters of the infants identified as having abnormal movement patterns were enrolled in multidisciplinary therapy services before 5 months of age, with improved continuity of follow-up. The GMA setup was described as having a high cost-to-benefit ratio.²² This kind of real-world implementation work matters enormously, because a screening tool that works beautifully in research but never reaches the babies who need it most is not solving the problem. The combination of smartphone-based video, growing clinician training programs, and emerging AI tools is steadily making GMA more accessible to families regardless of where they live.