What Is a UPDRS Score and What Does It Mean?

A UPDRS score is a standardized measure of how Parkinson’s disease affects someone across four domains: non-motor symptoms, motor symptoms in daily life, a clinician-performed motor exam, and medication-related complications. The scale, formally called the Movement Disorder Society-Unified Parkinson’s Disease Rating Scale (MDS-UPDRS), is the most widely used clinical tool for tracking Parkinson’s severity and is central to virtually every major drug trial for the disease. Higher scores mean worse symptoms, with each of the four parts scored separately and sometimes combined into a total. Understanding what these numbers actually tell you, and where they fall short, matters whether you’re a patient trying to make sense of a clinic visit or a caregiver watching scores change over time.

The Four Parts of the Scale

The MDS-UPDRS divides Parkinson’s disease into four distinct sections, each capturing a different aspect of how the disease affects a person’s life. This structure was designed so that clinicians and researchers could look at any one part independently or combine them for a fuller picture.

Part I, Non-Motor Experiences of Daily Living: Covers symptoms people don’t always associate with Parkinson’s, such as mood changes, sleep problems, pain, fatigue, cognitive difficulties, hallucinations, and urinary issues. Some items are answered by the patient or caregiver, while others are rated by the clinician. Validation studies have shown that Part I correlates strongly with other established non-motor symptom scales, confirming it captures this often-overlooked dimension of the disease well.¹
Part II, Motor Experiences of Daily Living: Asks how motor symptoms interfere with everyday activities like eating, dressing, writing, walking, and getting out of bed. This part is primarily patient-reported or completed with help from a caregiver.²
Part III, Motor Examination: The clinician-administered section and the most heavily used part in research. A neurologist rates specific motor signs including slowness of movement, tremor, rigidity, gait, balance, and speech during an in-office exam.³
Part IV, Motor Complications: Tracks problems that develop as a side effect of Parkinson’s medications over time, particularly dyskinesias (involuntary movements) and “off” periods when medication wears off and symptoms return.⁴

The original UPDRS was developed in the 1980s and became the standard in Parkinson’s research. The Movement Disorder Society organized a formal revision that resulted in the MDS-UPDRS, which retained the four-part structure but expanded it to better capture non-motor symptoms and resolved measurement problems identified in the original version.⁵

How Scoring Works

Every individual item on the MDS-UPDRS is rated on a five-point scale from 0 (normal) to 4 (severe). The items within each part are added up to give a part total. Part I has 13 items (maximum 52 points), Part II has 13 items (maximum 52 points), Part III has 18 items with some scored on each side of the body (maximum 132 points), and Part IV has 6 items (maximum 24 points). A total score across all four parts can reach 260, though in practice even patients with advanced disease rarely come close to that ceiling.

In clinical trials you’ll most often see Part III scores reported, since the motor examination is the most standardized and reproducible portion of the scale. Parts I and II rely on patient self-report, which adds a subjective element but also captures information a neurologist can’t directly observe during a brief clinic visit.

Proposed Severity Thresholds and Their Limits

Researchers have proposed cutoff scores to classify patients as mild, moderate, or severe within each part of the scale. One widely cited study defined the dividing lines as follows: Part I mild-to-moderate at around 10–11 points and moderate-to-severe at 21–22; Part II at 12–13 and 29–30; Part III at 32–33 and 58–59; and Part IV at 4–5 and 12–13.⁶

These cutoffs sound tidy, but there’s a real problem with taking them at face value. A total score is a sum, and two people can arrive at the same number through very different combinations of item ratings. Someone might score a 4 (severe) on two items and 0 on everything else, landing at an 8 for Part I, which falls below the “mild” threshold. But having two symptoms rated as severe is not the same as having mild disease. This is why experts have cautioned against treating these thresholds as firm diagnostic categories: the individual item scores matter as much as, or more than, the total.⁷

In practice, most neurologists use the score as one input among several. They look at which specific items are driving the total, how scores change over time, and how the numbers line up with what the patient describes. A single snapshot score without that context can be misleading.

What Counts as a Meaningful Change

When you or your doctor tracks MDS-UPDRS scores over months or years, small fluctuations are expected. The question is: how much does a score need to change before it reflects a real, noticeable difference in your symptoms? Researchers have calculated these thresholds, called the minimal clinically important difference, for each part of the scale.

For Part I (non-motor experiences), an improvement of about 2.6 points or a worsening of about 2.5 points is considered the smallest change likely to be clinically meaningful. For Part II (motor daily living), the thresholds are roughly 3 points in either direction. For Part III (motor exam), the thresholds are asymmetric: an improvement of about 3.25 points is meaningful, while a worsening needs to be closer to 4.6 points before it signals a real decline.⁸⁹ For Part IV (motor complications), the numbers are smaller, around 0.9 points for improvement and 0.8 for worsening.¹⁰

The asymmetry in Part III is worth pausing on. It takes a bigger worsening than improvement for the change to be considered clinically meaningful. One reason may be that patients and clinicians are more attuned to deterioration and may flag smaller declines as concerning, while improvements need to cross a higher bar to feel genuinely different. When combined parts are assessed together, the thresholds scale up: for Parts I through IV combined, roughly 7 points of improvement or 6 points of worsening is the meaningful threshold.

How Scores Track with Disease Stage

The Hoehn and Yahr scale is a separate, simpler staging system for Parkinson’s that ranks patients from stage 1 (symptoms on one side only) through stage 5 (wheelchair-bound or bedridden). MDS-UPDRS scores increase substantially through each of these stages. On average, Part III (motor exam) scores jump by about 14.6 points per stage, which is by far the steepest climb. Part II (motor daily living) increases by about 7.7 points per stage, Part I (non-motor) by about 3.8 points, and Part IV (motor complications) by about 2 points.¹¹

The takeaway is that the motor exam scores are the most sensitive to advancing disease stage. But Part I and Part II scores also rise consistently, which underscores that Parkinson’s progression is not purely about motor decline. Sleep problems, mood disturbances, and cognitive changes accumulate alongside the visible motor symptoms.

The Quality-of-Life Connection

Patients and caregivers understandably care less about a number on a rating scale and more about how the disease affects daily life. Research using the PDQ-39, a well-established quality-of-life questionnaire for Parkinson’s, has found that the non-motor and daily-living parts of the MDS-UPDRS (Parts I, II, and IV) are more closely linked to overall quality of life than the motor examination (Part III).¹² Pain, fatigue, and depressed mood were among the individual items most strongly associated with worse quality of life. A separate analysis from a phase 3 trial did find that Part III scores correlated with quality of life as well, particularly with the activities-of-daily-living domain.¹³

This matters because drug trials often use Part III as their primary endpoint, since it is the most standardized and objective measure. But a treatment that improves your motor exam score by a few points may not be the same one that makes you feel better day to day if your main burden is fatigue, pain, or anxiety. Patients who are given their Part III scores after a clinic visit should know those numbers capture one dimension of the disease, not the whole experience.

Day-to-Day Score Variability

One complication that surprises many patients is how much MDS-UPDRS Part III scores can fluctuate from one visit to the next, even without any real change in the underlying disease. A study that measured motor scores repeatedly in the same patients found that roughly 30% of the score changes between visits exceeded the threshold for clinically meaningful improvement, and about 17% exceeded it for worsening, simply due to natural day-to-day variability. The average size of these fluctuations was around 9 to 11 points in either direction. Factors you might expect to matter, like medication dose, stress, fatigue, sleep quality, and recent exercise, were not significantly associated with the variability.¹⁴

For patients, this means a single visit’s score might paint a picture that’s either rosier or gloomier than your typical day. If your neurologist sees a jump in your Part III score at one visit, it doesn’t necessarily mean your disease has progressed. This is one reason clinicians prefer to look at trends across multiple assessments rather than reacting to any single number.

Why Early Disease Is Hard to Measure

The MDS-UPDRS was designed to cover the full range of Parkinson’s severity, but evidence suggests it struggles at the mild end of the spectrum. An analysis of data from a large early-Parkinson’s cohort found that most items showed a strong floor effect in early disease, meaning many patients scored 0 (normal) on a large number of items, leaving little room for the scale to detect subtle progression.¹⁵ There was also a gap in the scale at very mild severity levels, meaning the instrument couldn’t distinguish between “just barely detectable” and “not yet detectable.”

Researchers have explored subscore approaches to get more sensitivity in early disease. One recent study found that focusing specifically on limb-related slowness and rigidity items produced a more finely graded severity measure: signs typically appeared first in the arm on the affected side, then the leg on that side, and finally the opposite limbs.¹⁶ This kind of subscoring may eventually improve how early-stage trials detect treatment effects, but it hasn’t replaced the standard scoring approach yet.

What Brain Imaging Adds to the Picture

Doctors sometimes order dopamine transporter imaging scans (DaT scans) to help confirm a Parkinson’s diagnosis or assess its biological severity. These scans measure how much dopamine-producing nerve activity remains in specific brain regions. Research has found that lower dopamine transporter binding in the putamen, a key motor region of the brain, correlates with higher MDS-UPDRS Part III scores.¹⁷ Longitudinal data tracking patients over four years has also shown that changes in dopamine transporter binding track with changes in MDS-UPDRS scores over time.¹⁸

The correlation is real but imperfect. The relationship between dopamine loss and clinical symptoms is not one-to-one, because the brain compensates for a long time before symptoms appear, and non-dopaminergic processes contribute to many Parkinson’s symptoms. Brain imaging gives a biological anchor for the clinical score, but neither measure tells the complete story on its own.

Wearable Sensors and Remote Monitoring

One of the biggest limitations of the MDS-UPDRS is that it captures a snapshot during a clinic visit, which may last 20 or 30 minutes and happen only a few times a year. Wearable devices, like accelerometers worn on the wrist, are being developed to provide continuous measurement of motor symptoms between visits. Algorithms trained on wearable data have shown strong correlations with clinician-rated motor scores: one wearable system’s bradykinesia measure correlated closely with UPDRS Part III motor scores, and its dyskinesia measure showed even tighter agreement with clinical ratings.¹⁹

A multi-sensor validation study in Japan found that composite scores from wearable devices correlated moderately with neurologist consensus ratings, with the overall motor composite achieving a correlation of 0.70 with Part III scores.²⁰ Machine learning models have also been used to estimate Part III scores from wearable data, though with meaningful error margins of roughly 10 points on average, which is larger than the clinically meaningful change threshold.²¹

These technologies are not ready to replace the clinical exam, but they are increasingly being used alongside it, particularly in clinical trials where capturing symptom fluctuations between visits could reveal treatment effects that a quarterly office visit would miss.

Reliability Across Languages and Cultures

Because Parkinson’s is a global disease, the MDS-UPDRS has been translated into many languages through a formal five-step process overseen by the International Parkinson and Movement Disorder Society. The process includes independent translation and back-translation, cognitive testing with native speakers, field testing with Parkinson’s patients, and full statistical validation against the English version.²²

The Japanese translation, for instance, met all pre-specified criteria to be designated an official translation, with statistical fit indices exceeding the minimum standard relative to the English version.²³ Some minor differences between language versions have been identified, but they fall within acceptable ranges. This matters for multinational clinical trials, where patients across countries need to be assessed on equivalent scales for the results to be comparable.

How the Score Is Used in Clinical Trials

If you follow Parkinson’s research, you’ll see MDS-UPDRS scores in nearly every trial result. Part III is the workhorse endpoint: when a press release says a new drug “significantly improved motor function,” it almost always means the drug group had a bigger drop in Part III scores than the placebo group. The minimal clinically important difference thresholds described earlier are the yardstick for whether the improvement is large enough to matter to patients, not just statistically significant in a spreadsheet.

Newer trial designs are also looking beyond Part III. Researchers have proposed tracking the emergence of new symptoms using Parts I and II as an outcome measure, which could capture whether a treatment slows the spread of the disease into new domains rather than just improving existing symptoms.²⁴ This is a shift from asking “did the score go down?” to asking “did the disease stay contained?” and it reflects growing recognition that the total score doesn’t capture everything patients care about.

When the Scale Doesn’t Apply

The MDS-UPDRS was developed for and validated in people with Parkinson’s disease. It is sometimes used in related conditions like multiple system atrophy or progressive supranuclear palsy, but the results should be interpreted carefully. In atypical parkinsonism conditions, the motor exam scores don’t correlate with quality of life the way they do in typical Parkinson’s.²⁵ The disease mechanisms and symptom profiles differ enough that the scale can miss what matters most in those conditions, and non-motor symptom scales may be more relevant for tracking burden.

Part IV (motor complications) also has limitations within Parkinson’s itself. A validation study found that the individual items for detecting “off” time and dyskinesias had sensitivities and specificities in the range of 67–74%, meaning they miss a meaningful fraction of complications that patients actually experience.²⁶ A substantial floor effect has also been noted for Part IV, with about 30% of patients scoring zero, partly because motor complications tend to develop years into the disease after extended medication use.²⁷ If you’re early in your Parkinson’s course and haven’t developed these complications yet, your Part IV score will be zero, and that’s expected rather than informative.