What Kinds of Tests Are Done for Headaches?

Most headaches are diagnosed through a careful clinical history and neurological exam, not through scans or lab work. When your doctor asks about timing, location, severity, and associated symptoms, that conversation is itself the most important diagnostic tool. Testing becomes necessary only when something in the clinical picture raises concern about a secondary cause, such as a brain bleed, infection, or elevated pressure inside the skull. The range of tests available is broad, from brain imaging and blood draws to spinal taps and eye exams, and the decision about which to order depends almost entirely on which red flags are present.

When Testing Is and Isn’t Needed

The American Headache Society has been clear on this point: there is no need for brain imaging in patients whose headaches fit a migraine pattern, whose neurological exam is normal, and who have no unusual features or warning signs.¹ That guidance applies to the vast majority of people with headaches. Tension-type headaches, migraines, and cluster headaches are all “primary” headaches, meaning the headache itself is the condition rather than a symptom of something else. For these, the diagnosis rests on your description of symptoms and a normal neurological check.

The challenge is distinguishing primary headaches from “secondary” ones, where the headache signals an underlying problem. A structured screening tool called the SNNOOP10 list helps clinicians systematically check for warning signs that point toward secondary causes.² The name is a mnemonic, and it covers things like systemic symptoms (fever, weight loss), neurological deficits, sudden onset, onset after age 50, a change in an established headache pattern, headache triggered by coughing or exertion, pregnancy, eye pain with autonomic features, history of head trauma, and use of medications or substances that could cause headaches. If any of these red flags are present, testing moves from optional to important.

Brain Imaging

When imaging is warranted, the two main options are CT (computed tomography) and MRI (magnetic resonance imaging). They serve different purposes and are not interchangeable.

CT scans are fast, widely available, and particularly good at detecting bleeding. For a thunderclap headache, the sudden, explosive kind that reaches maximum intensity within seconds, a non-contrast CT scan done within six hours has a sensitivity around 99% for detecting subarachnoid hemorrhage, a type of brain bleed that can be life-threatening.³ That speed and accuracy make CT the go-to test in emergency settings when a bleed is suspected. The downside is radiation exposure, which matters more in children and in situations where repeated scans are needed.

MRI provides more detail about brain tissue, blood vessels, and structures around the brain. The European Headache Federation recommends brain MRI along with MR angiography and MR venography for specific migraine complications like persistent aura without infarction or migrainous infarction, where blood vessel imaging is critical.⁴ MRI with MR venography can also help identify idiopathic intracranial hypertension, a condition where pressure inside the skull is abnormally high, causing headaches and threatening vision.⁵ MRI takes longer than CT, costs more, and isn’t always available after hours, which is why it’s typically reserved for non-emergency situations or when CT results are inconclusive.

CT Angiography

CT angiography (CTA) adds a contrast dye to a standard CT scan to visualize blood vessels directly. It’s especially useful when a brain aneurysm is suspected. In one study of over 500 patients with acute severe headache who underwent CTA, about 7% had intracranial vascular abnormalities, including aneurysms, arterial dissection, and moyamoya disease.⁶ Research has found that combining a standard CT with CTA can rule out aneurysmal subarachnoid hemorrhage with better than 99% confidence.⁷ This combination has shifted practice in many emergency departments, sometimes reducing the need for lumbar puncture as a follow-up test.

Lumbar Puncture and Spinal Fluid Analysis

A lumbar puncture, commonly called a spinal tap, involves inserting a needle into the lower back to collect cerebrospinal fluid (CSF). It serves two purposes in headache evaluation: measuring the pressure of the fluid and analyzing its contents.

Measuring opening pressure is essential when idiopathic intracranial hypertension is suspected. In that condition, elevated CSF pressure causes headaches and can damage the optic nerves. One study found that lumbar puncture opening pressure in these patients averaged around 32 cmCSF.⁸ Interestingly, the pressure reading doesn’t always predict how bad the headache is. On the other end of the spectrum, spontaneous intracranial hypotension causes headaches from low CSF pressure, but only about a third of those patients actually have pressure readings low enough to meet the diagnostic cutoff. Most fall in the low-normal to normal range.⁹ This means a normal pressure reading doesn’t always rule out a pressure-related headache.

CSF analysis is also used to check for subarachnoid hemorrhage when a CT scan comes back normal but clinical suspicion remains high. The fluid is examined for blood products and for xanthochromia, a yellowish discoloration caused by the breakdown of red blood cells. Xanthochromia detected by spectrophotometry is highly sensitive and specific for subarachnoid hemorrhage.¹⁰ In a large observational study of over 2,200 patients who received lumbar punctures after a normal head CT, about 5% had CSF positive for blood suggesting hemorrhage, though on further imaging only a small fraction of those turned out to have aneurysms.¹¹ The challenge with lumbar puncture is distinguishing a true positive from a “traumatic tap,” where the needle itself causes bleeding that contaminates the sample. Statistical analysis of the fluid can help: bloody CSF and xanthochromic supernatant are strong predictors of true hemorrhage rather than needle artifact.¹²

Blood Tests

Routine blood work doesn’t diagnose most headaches, but specific blood tests become critical in certain clinical scenarios. The most important example involves giant cell arteritis, also known as temporal arteritis, a condition that causes inflammation in blood vessels (usually around the temples) and can lead to sudden, permanent vision loss if untreated.

When giant cell arteritis is suspected, typically in someone over 50 with a new severe headache, scalp tenderness, or jaw pain while chewing, the first-line tests are the erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP). Both measure inflammation in the body. In studies comparing the two, CRP tends to be slightly more sensitive, catching about 87% of biopsy-confirmed cases compared to about 84% for ESR.¹³ The traditional diagnostic criterion uses an ESR above 50 mm/h, and in one study, 69% of patients with biopsy-proven disease had levels that high.¹⁴

Here’s the catch: a normal ESR does not rule out the disease. Case reports describe patients with clear clinical symptoms of giant cell arteritis who had ESR values well below the 50 mm/h threshold.¹⁵ About 4% of patients with biopsy-confirmed disease had both a normal ESR and a normal CRP at the time of diagnosis.¹³ If the clinical picture is suspicious, treatment typically starts immediately, even before blood results come back, because the risk of blindness is too high to wait.

Other blood tests that can matter in headache workups, depending on context, include thyroid function tests, a complete blood count to check for infection or anemia, and metabolic panels when medication-related headaches or systemic illness are suspected. These are less headache-specific and more about ruling out general medical conditions that can manifest as head pain.

Temporal Artery Biopsy

When blood tests and clinical symptoms point toward giant cell arteritis, a temporal artery biopsy often follows. This is a minor surgical procedure where a small segment of the temporal artery is removed and examined under a microscope for the inflammatory changes characteristic of the disease. It remains the gold standard for confirming the diagnosis.

In a review of nearly 500 biopsies, about 13% came back positive. The strongest predictors of a positive result were jaw or tongue pain during chewing, elevated inflammatory markers, and older age. Using a risk-stratification approach, only about 3% of low-risk patients tested positive compared to roughly 44% of high-risk patients.¹⁶ Earlier research showed that patients with positive biopsies were much more likely to have headache (93% vs 62%) and jaw claudication (50% vs 18%) compared to those with negative biopsies, though no single symptom was reliable enough on its own to confirm or exclude the diagnosis.¹⁷

Eye Exams and Optical Coherence Tomography

Looking at the back of the eye can reveal swelling of the optic nerve head, known as papilledema, which signals elevated intracranial pressure. Traditionally this required dilating the pupils and having an experienced clinician peer through an ophthalmoscope. Newer technology is changing how and where this check happens.

Nonmydriatic fundus photography combined with optical coherence tomography (OCT) allows emergency departments to image the back of the eye without dilating drops. In one study of 129 consecutive headache patients in an emergency department, this approach identified papilledema in about 4% and other relevant findings in another 6%, enabling rapid decisions about whether further testing was needed.¹⁸ OCT is also valuable for monitoring idiopathic intracranial hypertension over time. Research has shown that patients with this condition have increased optic nerve head volume on OCT, and that these measurements correlate with CSF pressure, making it a useful non-invasive way to track the disease.¹⁹

Nerve Blocks as a Diagnostic Tool

Not all headache tests involve scans or needles in the spine. For headaches suspected to originate from the neck or the occipital nerves at the back of the skull, a diagnostic nerve block can provide both answers and relief. In cervicogenic headache and occipital neuralgia, injecting an anesthetic around the greater or lesser occipital nerve can tell the clinician whether that nerve is the pain source. If the headache disappears temporarily after the block, it confirms the diagnosis and opens the door to longer-term treatments.²⁰ This approach is particularly useful for headaches that don’t fit neatly into migraine or tension-type categories and that seem connected to neck movement or tenderness at the base of the skull.

EEG Testing

Electroencephalography, which measures electrical activity in the brain, is not a standard headache test. It’s occasionally ordered when a clinician suspects that headache episodes could be related to seizure activity, especially in children. A multi-center study of pediatric headache patients found that about a quarter of children with migraine with aura had abnormal EEG readings, compared to roughly 11% of those with migraine without aura.²¹ The overlap between migraine aura and epilepsy can make clinical diagnosis tricky, and EEG helps sort that out. For adults with straightforward headache patterns, EEG rarely adds useful information.

The Sinus Headache Problem

One of the most common testing dead ends involves sinus headaches. Many people who believe they have sinus headaches actually have migraines. The confusion is understandable: migraines often involve forehead and cheek pain, nasal congestion, and watery eyes, all of which overlap with sinusitis symptoms.²² This leads to sinus CT scans, ENT referrals, and sometimes unnecessary treatments. One study found that when patients with self-described “sinus headaches” who had no evidence of sinusitis on endoscopy and CT scan were given migraine-specific medication (triptans), 82% had a significant response, supporting the idea that these headaches were migraines all along.²³

Complicating things further, even genuine migraine patients sometimes do have abnormal sinus findings on CT. One study found that many patients meeting migraine criteria also had radiographic evidence of sinus disease, raising questions about whether the two conditions simply coexist or whether the abnormal sinus findings are incidental.²⁴ The presence or absence of purulent nasal discharge is the most reliable clinical clue for telling the two apart. If you’ve been treated repeatedly for “sinus headaches” without lasting improvement, a migraine evaluation may be more productive than another scan.

Headache Testing in Children

The testing approach for children with headaches is even more conservative than for adults. Most neuroimaging studies performed on children with headaches come back normal, and the scans themselves carry costs beyond the financial, including radiation exposure from CT and the need for sedation in younger children who can’t hold still in an MRI scanner.²⁵ European neuroradiology guidelines emphasize that clinical history and physical examination should drive the initial evaluation, and that imaging should be avoided entirely in children with primary headaches and low-risk mild head injuries. When imaging is necessary, techniques that avoid radiation and sedation should be prioritized.²⁶

The Cost of Overtesting

Despite guidelines recommending against routine imaging for typical headaches, many scans are still ordered. An analysis of neuroimaging expenditures found that total spending reached an estimated $3.9 billion over a four-year period, with $1.5 billion of that coming from migraine-related visits alone.²⁷ Much of this spending produces normal results that don’t change treatment. Overtesting also creates its own problems: incidental findings on brain MRI, like small cysts or benign abnormalities, can trigger anxiety and lead to further unnecessary testing. The evidence consistently points toward clinical judgment and structured red-flag screening as the best gatekeepers for when imaging is truly needed.

Computerized Diagnostic Tools

A newer development in headache diagnosis is the use of computerized tools and machine-learning algorithms to help classify headache types. A systematic review of these tools found a median diagnostic accuracy of about 89%, with sensitivity around 87% and specificity around 90%.²⁸ Half of the tools studied were self-administered, meaning patients could fill them out before seeing a clinician. Most were evaluated in specialized headache clinics rather than general practice settings, so how well they perform in a typical doctor’s office remains an open question. These tools are not replacements for clinical evaluation, but they could eventually help reduce diagnostic delays, which can stretch for years, especially for less common headache types. One case report documented an eleven-year delay in diagnosing cluster headache, with the patient initially misdiagnosed with migraine and then with an ear problem, before a detailed reassessment of the clinical history led to the correct diagnosis.²⁹ If a screening questionnaire could flag cluster headache features earlier, years of incorrect treatment might be avoided.