Most panels that screen for inflammation now lean on one number: hs-CRP, the marker we covered in an earlier piece on the quiet fire you cannot feel. But CRP was not the first inflammation test, and it did not fully retire the one that came before it. Erythrocyte sedimentation rate, ESR, is over a century old, measures inflammation by a strange indirect method, and still earns a place on specific diagnostic checklists that CRP cannot fill. Here is what the test actually does, why CRP pushed it out of routine screening, and where it still refuses to go away.
What ESR actually measures
CRP is a direct chemical reading: a liver-made protein, produced in response to the signaling molecule interleukin-6, measured in the blood. ESR measures nothing chemical at all. It is a physical race. A tube of anticoagulated blood is left standing upright, and the test simply clocks how far the red blood cells fall in an hour.
In a healthy sample, red blood cells carry a slight negative surface charge that keeps them apart, so they settle slowly. Inflammation changes the plasma around them. Acute-phase proteins, fibrinogen most of all, coat the red cells and reduce that charge, letting them stick together into stacked clusters called rouleaux. Clustered cells are heavier for their surface area and fall faster, so a higher ESR reading is really a read-out of how much fibrinogen and other proteins have accumulated in the plasma, not a direct count of inflammation itself.
The test's age shows in its method. It was first described in the 1890s by a physician studying rheumatic fever, who noticed elevated fibrinogen made blood settle faster, and it was standardized in the 1920s using sodium citrate as the anticoagulant, the version still referenced today as the Westergren method. Newer instruments can return a result through centrifugation in minutes instead of an hour, but the underlying principle has not changed in a hundred years.
Why CRP took over
CRP and ESR are frequently ordered together, but they run on different clocks, and that difference is why CRP became the default. CRP rises within about four to six hours of an inflammatory trigger, peaks at roughly one to two days, and has a half-life of about 19 hours, climbing and falling quickly enough to track an active process in near real time. Fibrinogen, the protein ESR depends on, rises more slowly over 24 to 48 hours and can stay elevated for weeks after the original event has resolved, which makes ESR a laggard for catching acute change and a poor tool for confirming a flare has actually settled.
The College of American Pathologists' own guidance on the two tests is direct about the tradeoff: CRP is recommended over ESR to detect acute-phase inflammation in patients with an undiagnosed condition, since it responds faster and reflects the acute-phase response more specifically. A 2025 review from the clinical laboratory professional group ADLM makes a similar case for acute care, noting that ESR requires its own separate instrument, adds complexity to a lab's test menu, and offers little that a same-day CRP does not already cover. That case has moved practice: a 2024 study in Hospital Pediatrics is one of several papers describing formal hospital programs built specifically to reduce routine ESR ordering that was not changing what a clinician did next.
Where ESR still holds its ground
None of that has made ESR obsolete. It remains the screening test of choice for two specific conditions: giant cell arteritis, sometimes called temporal arteritis, and the closely related polymyalgia rheumatica. In both, ESR is classically markedly elevated, often above 50 to 100 mm per hour, and that magnitude is written directly into the diagnostic criteria clinicians use. Its slow kinetics, the same trait that makes it a weak acute-infection screen, work in its favor here, because both conditions are sustained and chronic rather than sharp and self-limiting.
ESR also earns a seat in a few other specific corners. It helps monitor multiple myeloma, where an abnormal protein called paraprotein coats red cells and raises the sedimentation rate through a mechanism that has nothing to do with cytokine-driven inflammation at all. And it can pick up occult, slow-building infections, such as subacute osteomyelitis or bacterial endocarditis, in cases where the acute CRP spike may have already faded by the time a patient is worked up.
Reading the two tests together, rather than picking one, is where the real information lives. Both markers elevated together points toward active, sustained inflammation. A high ESR with a normal CRP is a more specific pattern: it suggests a chronic or resolving process, high fibrinogen on its own, anemia (which lowers the red cells' natural charge-repulsion and speeds settling independent of any inflammation), or a paraprotein state. Both normal makes systemic inflammation less likely, though it does not rule it out, since some inflammatory conditions run with a limited marker response.
The evidence still being written
The newest thread in this story circles back to the protein ESR was always indirectly measuring. A pilot study out of Ireland tested whether measuring fibrinogen directly, rather than inferring its effect through a sedimentation tube, could track disease activity in polymyalgia rheumatica at least as well as CRP and ESR combined, and found it held up as accurate and, notably, more specific for confirming a real treatment response. It is a small study, not yet the kind that changes a guideline, but a fitting turn for a hundred-year-old indirect test: the field is now asking whether measuring the direct cause was the better idea all along, just one current instruments made hard to do cheaply until recently.
The honest summary is not that one test beat the other. CRP is quicker, more specific to inflammation itself, and the better default when a clinician is asking whether something acute is happening right now. ESR is slower, cheaper worldwide, non-specific by nature, and still unmatched for a small number of chronic and rheumatologic conditions where its very sluggishness is the point. Neither result stands alone as a diagnosis; both are read against symptoms, exam findings, and a clinical question, not as a number to chase down on its own.
Common questions
Is a normal ESR enough to rule out inflammation? No. ESR is nonspecific and can be affected by age, sex, anemia, kidney function, pregnancy, and protein levels that have nothing to do with inflammation, and some inflammatory conditions produce only a limited rise. A normal result lowers the likelihood of an active inflammatory process but does not exclude one, especially early or localized disease.
Why would a doctor order both CRP and ESR instead of just one? They track different things on different timelines. CRP reflects an acute chemical signal that rises and falls within days; ESR reflects an indirect, slower-moving physical effect that can stay elevated for weeks, and its reference range also runs higher with age. Ordering both, in the right clinical context, can distinguish a fresh flare from a chronic or resolving one, since the pattern between the two carries more information than either number alone.
What is rouleaux formation? It is the stacking of red blood cells into coin-like clusters, driven mainly by fibrinogen and other acute-phase proteins coating the cells and reducing the natural charge that normally keeps them apart. Clustered cells fall faster in the ESR tube, which is the entire physical basis of the test.
Is fibrinogen itself ever tested directly? It can be, and a small pilot study in polymyalgia rheumatica found direct fibrinogen measurement performed at least as well as CRP and ESR for tracking disease activity, with some evidence it was more specific for confirming a genuine treatment response. It is not yet a routine substitute for either standard test.
Why is ESR sometimes still used when CRP is faster and more specific? Its slow, sustained kinetics are actually the advantage in chronic, smoldering conditions like giant cell arteritis and polymyalgia rheumatica, where a marker that stays elevated over weeks is more diagnostically useful than one built to spike and fade within days.
