Scores the neurological deficit of an acute stroke from 15 bedside items, from 0 (no deficit on the scale) to 42.
Result
Score every item to see the total.
0 of 15 items scored
Sources
ESO–ESMINT 2024: Weak for intervention · 7 studies
Weak for interventionESO–ESMINT 2024 suggests thrombectomy for basilar occlusion when the NIHSS is 10 or more.
7
studies
5,927
citations
17,088
PubMed papers
Evidence
Guideline recommendation
“For patients presenting with severe symptoms (NIHSS ⩾ 10), we suggest BMT + EVT over BMT only.”
European Stroke Organisation and ESMINT, 2024: basilar artery occlusion
Weak for intervention, very low quality of evidence
Eur Stroke J · cited 49 times
PubMed · Free full text · DOI
Known limitations
The 15-item scale was designed for acute stroke trials. In 24 patients, interrater reliability was high (mean kappa 0.69) and test–retest kappa 0.66 to 0.77, with no difference between a neurologist, a resident and nurses. In 65 patients the score correlated with infarct size on CT at 1 week (r = 0.68) and with outcome at 3 months (r = 0.79).
Four stroke fellows rated 20 patients in pairs. Agreement between observers was moderate to substantial for 9 of the 13 items tested.
After video training, 162 investigators rated 11 taped patients. Most items had moderate to excellent agreement (unweighted kappa above 0.60), but facial paresis and ataxia had poor agreement (below 0.40). The method became the basis of NIHSS certification.
In 1,281 patients from the TOAST trial, each additional point lowered the chance of an excellent outcome by 24% at 7 days and 17% at 3 months. A score of 16 or more forecast a high probability of death or severe disability, and 6 or less a good recovery.
In 2,152 patients with vessel imaging, 48.5% had an occlusion. Within 3 hours, a score of 9 or more predicted an anterior circulation occlusion with a positive predictive value of 86.4%; only 5% of central occlusions scored below 4. The association was poor in the posterior circulation and after 6 hours.
In the placebo arm of the NINDS t-PA trial, the median volume of right hemisphere infarcts was about double that of left hemisphere infarcts in each 5-point band below 20. At a 24-hour score of 16 to 20 the medians were 133 mL against 48 mL.
Twenty patients with persistent symptoms and an infarct on diffusion-weighted MRI had an NIHSS of 0. Headache, vertigo and nausea were common, the posterior circulation was often affected, and truncal ataxia was the most frequent sign.
Ann Emerg Med · cited 153 times
PubMed · Free full text · DOI
The original paper has been cited 5,927 times, and 17,088 PubMed articles mention the NIH Stroke Scale in their title or abstract. Citation counts from OpenAlex and article counts from PubMed, retrieved November 6, 2026. Journal Metrics Scores and quartiles are from this site’s current release; follow the journal name for details.
After scoring
Advice · Acute stroke · Critical actions
Record the time, and score again when the patient changes. Use the score with imaging and the time last known well.
Next steps
Avoid common errors
6 common errors to avoid
and 4 more
Before you record the score
1Score what the patient does
Record what the patient does, not what you think they could do, and go through the items in order. Do not give cues except where an item allows them (pantomime for 1c and the motor items); for 1b and 1c only the first attempt counts.
2Posterior circulation strokes score low
Vertigo, gait and truncal ataxia, nausea and swallowing problems are not measured, so brainstem and cerebellar strokes can score low, even 0 (Martin-Schild 2011). The score also predicts occlusion poorly in the posterior circulation (Heldner 2013).
3Right hemisphere strokes score lower
Seven points relate to language and two to neglect, so for the same score a right hemisphere infarct is often about twice as large as a left one (median volumes, Woo 1999).
4Ataxia only out of proportion to weakness
Do not score ataxia in a weak limb unless it is clearly more than the weakness explains, and score it 0 if the patient cannot understand or is paralyzed. Ataxia and facial palsy are the items examiners agree on least (Lyden 1994).
5Untestable items
UN is only for amputation or joint fusion (motor and ataxia items) and intubation or another physical barrier (dysarthria). It adds no points, so write down which items were untestable and why.
6Coma and aphasia rules
With 1a scored 3, sensory scores 2 and language 3. An aphasic or stuporous patient who cannot understand the questions scores 2 on item 1b; one who cannot speak for another reason, such as intubation, scores 1.
Formula
Fifteen items, each from 0 (normal) up to 2, 3 or 4 points, added for a total of 0 to 42. Because ataxia is not scored in a paralyzed limb or an unresponsive patient, real totals stop a little below 42.
15 items, 0 to 42 points
Fifteen items, each from 0 (normal) up to 2, 3 or 4 points, added for a total of 0 to 42. Because ataxia is not scored in a paralyzed limb or an unresponsive patient, real totals stop a little below 42.
Items and points
| Item | Finding | Points |
|---|---|---|
| 1a. Level of consciousness | Alert | 0 |
| Not alert, but arousable by minor stimulation | 1 | |
| Not alert, needs repeated or strong stimulation | 2 | |
| Unresponsive, or reflex responses only | 3 | |
| 1b. Questions: the month and their age | Both answers correct | 0 |
| One answer correct | 1 | |
| Neither answer correct | 2 | |
| 1c. Commands: open and close the eyes, grip and release | Performs both tasks correctly | 0 |
| Performs one task correctly | 1 | |
| Performs neither task | 2 | |
| 2. Best gaze | Normal | 0 |
| Partial gaze palsy | 1 | |
| Forced deviation or total gaze paresis | 2 | |
| 3. Visual fields | No visual loss | 0 |
| Partial hemianopia | 1 | |
| Complete hemianopia | 2 | |
| Bilateral hemianopia | 3 | |
| 4. Facial palsy | Normal symmetrical movements | 0 |
| Minor paralysis | 1 | |
| Partial paralysis | 2 | |
| Complete paralysis of one or both sides | 3 | |
| 5a. Motor, left arm | No drift | 0 |
| Drift | 1 | |
| Some effort against gravity | 2 | |
| No effort against gravity | 3 | |
| No movement | 4 | |
| Untestable (UN) | UN | |
| 5b. Motor, right arm | No drift | 0 |
| Drift | 1 | |
| Some effort against gravity | 2 | |
| No effort against gravity | 3 | |
| No movement | 4 | |
| Untestable (UN) | UN | |
| 6a. Motor, left leg | No drift | 0 |
| Drift | 1 | |
| Some effort against gravity | 2 | |
| No effort against gravity | 3 | |
| No movement | 4 | |
| Untestable (UN) | UN | |
| 6b. Motor, right leg | No drift | 0 |
| Drift | 1 | |
| Some effort against gravity | 2 | |
| No effort against gravity | 3 | |
| No movement | 4 | |
| Untestable (UN) | UN | |
| 7. Limb ataxia | Absent | 0 |
| Present in one limb | 1 | |
| Present in two limbs | 2 | |
| Untestable (UN) | UN | |
| 8. Sensory | Normal, no sensory loss | 0 |
| Mild to moderate loss | 1 | |
| Severe to total loss | 2 | |
| 9. Best language | No aphasia, normal | 0 |
| Mild to moderate aphasia | 1 | |
| Severe aphasia | 2 | |
| Mute, global aphasia | 3 | |
| 10. Dysarthria | Normal | 0 |
| Mild to moderate | 1 | |
| Severe | 2 | |
| Untestable (UN) | UN | |
| 11. Extinction and inattention | No abnormality | 0 |
| Inattention or extinction in one modality | 1 | |
| Profound hemi-inattention, or extinction in more than one modality | 2 |
| 0 | No deficit measured by the scale |
|---|---|
| 1–4 | Minor stroke |
| 5–15 | Moderate stroke |
| 16–20 | Moderate to severe stroke |
| 21–42 | Severe stroke |
Each limb separately
Untestable (UN)
Severity bands
Test cases
3 published cases, each also a test case
3 published cases; the calculator reproduces every score.
Worked examples
A 32-year-old woman with a right occipital and thalamic infarct, a left homonymous hemianopia on confrontation testing and mildly decreased sensation on the left; the rest of the examination was normal. The report gives an NIHSS of 3.
“NIHSS on admission was 3 (2 points for visual field defect, 1 point for sensory loss).”
The report scores the hemianopia 2, the score for a complete hemianopia. Items the report does not give points for were entered as 0.
Entered as: 1a. Level of consciousness: Alert (0); 1b. Questions: the month and their age: Both answers correct (0); 1c. Commands: open and close the eyes, grip and release: Performs both tasks correctly (0); 2. Best gaze: Normal (0); 3. Visual fields: Complete hemianopia (2); 4. Facial palsy: Normal symmetrical movements (0); 5a. Motor, left arm: No drift (0); 5b. Motor, right arm: No drift (0); 6a. Motor, left leg: No drift (0); 6b. Motor, right leg: No drift (0); 7. Limb ataxia: Absent (0); 8. Sensory: Mild to moderate loss (1); 9. Best language: No aphasia, normal (0); 10. Dysarthria: Normal (0); 11. Extinction and inattention: No abnormality (0).
Source: Chen CX et al., Front Genet 2026 (PubMed · Free full text · DOI)
A 54-year-old man with a right paramedian pontine infarct, left facial droop, drift of the left arm and leg, and mild to moderate dysarthria. The report lists every item on admission, with a total of 7.
Taken item by item from Table 1 of the report: partial facial paralysis (2), drift in the left arm (1) and leg (1), ataxia in two limbs (2) and mild to moderate dysarthria (1); every other item normal.
Entered as: 1a. Level of consciousness: Alert (0); 1b. Questions: the month and their age: Both answers correct (0); 1c. Commands: open and close the eyes, grip and release: Performs both tasks correctly (0); 2. Best gaze: Normal (0); 3. Visual fields: No visual loss (0); 4. Facial palsy: Partial paralysis (2); 5a. Motor, left arm: Drift (1); 5b. Motor, right arm: No drift (0); 6a. Motor, left leg: Drift (1); 6b. Motor, right leg: No drift (0); 7. Limb ataxia: Present in two limbs (2); 8. Sensory: Normal, no sensory loss (0); 9. Best language: No aphasia, normal (0); 10. Dysarthria: Mild to moderate (1); 11. Extinction and inattention: No abnormality (0).
Source: Manongi N et al., J Investig Med High Impact Case Rep 2023 (PubMed · Free full text · DOI)
A 54-year-old man with infective endocarditis, 15 hours of left-sided weakness, a shallow left nasolabial fold, grade III strength in the left limbs and reduced sensation on the left. The report gives an NIHSS of 8.
“NIHSS score was 8 points (facial palsy 2 points, left upper limb 2 points, left lower limb 2 points, sensation 1 point, dysarthria 1 point).”
Item points are the authors’. The report scores facial palsy 2 although it describes only a shallow nasolabial fold, which the form scores 1 (minor paralysis). Items the report does not give points for were entered as 0.
Entered as: 1a. Level of consciousness: Alert (0); 1b. Questions: the month and their age: Both answers correct (0); 1c. Commands: open and close the eyes, grip and release: Performs both tasks correctly (0); 2. Best gaze: Normal (0); 3. Visual fields: No visual loss (0); 4. Facial palsy: Partial paralysis (2); 5a. Motor, left arm: Some effort against gravity (2); 5b. Motor, right arm: No drift (0); 6a. Motor, left leg: Some effort against gravity (2); 6b. Motor, right leg: No drift (0); 7. Limb ataxia: Absent (0); 8. Sensory: Mild to moderate loss (1); 9. Best language: No aphasia, normal (0); 10. Dysarthria: Mild to moderate (1); 11. Extinction and inattention: No abnormality (0).
Source: Li R et al., Cureus 2025 (PubMed · Free full text · DOI)
FAQ
10 questions and answers
and 8 more answers
Questions
The National Institutes of Health Stroke Scale (NIHSS) is a 15-item bedside examination that measures the neurological deficit caused by a stroke. Each item is scored from 0 (normal) upward and the items are added for a total of 0 to 42. Brott and colleagues at the University of Cincinnati first published it in Stroke in 1989 for use in acute stroke trials. The version used today, with each limb scored separately, is the NINDS form adopted for the NINDS t-PA trial, and it is now the standard severity score in stroke care and research.
Go through the items in order: level of consciousness (1a), two questions (1b), two commands (1c), gaze, visual fields, facial palsy, motor function of each arm and each leg, limb ataxia, sensation, language, dysarthria, and extinction and inattention. Score what the patient does on the first attempt without coaching, then add the points. Items that cannot be tested because of an amputation, joint fusion or intubation are recorded as UN and add no points.
A grouping used in many studies is 0 for no deficit on the scale, 1 to 4 minor stroke, 5 to 15 moderate stroke, 16 to 20 moderate to severe stroke and 21 to 42 severe stroke. These bands are a convention, not part of the original scale, and a low score can still mean a disabling deficit such as aphasia or hemianopia.
Yes, as a group average. In 1,281 patients from the TOAST trial, each additional point lowered the chance of an excellent outcome by 24% at 7 days and 17% at 3 months; 46% of patients with scores of 7 to 10 and 23% with scores of 11 to 15 had an excellent outcome at 3 months. The authors concluded that a score of 16 or more forecasts a high probability of death or severe disability, and 6 or less a good recovery (Adams 1999).
Higher scores make an occlusion more likely, but there is no safe cut-off. In 2,152 patients with vessel imaging, a score of 9 or more within 3 hours of onset had a positive predictive value of 86.4% for an occlusion in the anterior circulation, and 7 or more between 3 and 6 hours 84.4%. Only 5% of patients with a central occlusion seen within 3 hours scored below 4. In the posterior circulation and after 6 hours, the score predicted occlusion poorly (Heldner 2013).
No. In a series of 20 patients with an infarct on diffusion-weighted MRI and an NIHSS of 0, the usual symptoms were headache, vertigo and nausea, the infarct was often in the posterior circulation, and truncal ataxia was the most common sign (Martin-Schild 2011). The scale does not test gait, truncal ataxia or swallowing.
For an intubated patient, item 1b scores 1 if they cannot speak for a reason other than aphasia, item 9 is tested by writing, and dysarthria (item 10) is recorded as UN. When item 1a is 3 (unresponsive), the instructions give 2 for sensory loss (item 8) and 3 for language (item 9), and ataxia (item 7) is absent because the patient cannot understand the task.
Not quite. Seven of the 42 points relate to language and only two to neglect, so right hemisphere strokes score lower for the same size of infarct. In the NINDS t-PA trial, the median infarct volume of right hemisphere strokes was about double that of left hemisphere strokes in each 5-point band below 20 (Woo 1999).
Reliability is good for most items when examiners are trained. In the original 1989 version of the scale the mean interrater kappa was 0.69 (Brott 1989). After video training of 162 investigators, most items had moderate to excellent agreement, but facial palsy and ataxia had poor agreement (kappa below 0.40) (Lyden 1994). This video training became the basis of NIHSS certification.
No. This calculator scores the full 15-item NIHSS. Shortened versions, such as the modified NIHSS, leave out some items and give lower totals, so the two should not be mixed in one record or compared directly.
Further reading
13 references with PubMed and DOI links
13
references
6
free full text
From Brott T et al., Stroke 1989 to the newest case reports.
References
Contributed by Dr. Meng Zhao (MD · Former Neurosurgeon · Medical AI Researcher), from the original publications listed here. Last updated November 6, 2026. LinkedIn
The calculator code is tested against every worked example on this page before each release.
This calculator supports, and does not replace, clinical judgment. It is meant for healthcare professionals and is not medical advice for patients. Report an error through the contact page.
clinical grade of aneurysmal subarachnoid hemorrhage, I to V.
total calcium adjusted for albumin, in mg/dL or mmol/L.
The journal that published the original scale in 1989, with its current metrics.
Compare journals before you submit a stroke paper.