Korotkoff sounds are the series of thumping, whooshing, and tapping noises you hear through a stethoscope when a blood pressure cuff slowly deflates around someone’s upper arm. They are the foundation of manual blood pressure measurement, and they arise from turbulent blood forcing its way through a partially squeezed artery. A Russian surgeon named Nikolai Korotkoff first described them in 1905, and clinicians have since divided them into five distinct phases, each marking a different stage of blood flow as cuff pressure drops from above systolic to below diastolic.
How Blood Flow Creates the Sounds
When a blood pressure cuff is inflated high enough, it completely flattens the brachial artery in your upper arm. No blood passes through, and the stethoscope picks up silence. As cuff pressure is gradually released, there comes a moment when the peak pressure of a heartbeat is just strong enough to force the artery open briefly. Blood surges through that narrow opening in a turbulent jet, and that turbulence vibrates the vessel wall, producing an audible sound. The principle is the same as the noise you hear when you partially kink a garden hose: restrict the flow path and the smooth stream becomes noisy.
As the cuff continues to deflate, the artery stays open for a longer portion of each heartbeat, the turbulence changes character, and eventually the artery is no longer compressed at all. At that point, blood flows smoothly again, turbulence disappears, and the sounds stop. The entire sequence, from first sound to silence, maps the transition from full arterial compression to free flow, and each phase along the way has a recognizable acoustic signature.
Where the Discovery Came From
In November 1905, Nikolai Korotkoff presented a short report at the Imperial Military Medical Academy in St. Petersburg describing how, after fully compressing the arm with a Riva-Rocci cuff and then slowly lowering the pressure, a series of distinct sounds could be heard through a stethoscope placed over the artery below the cuff.1PubMed. A centenary of auscultatory blood pressure measurement: a tribute to Nikolai Korotkoff Before Korotkoff’s technique, measuring blood pressure non-invasively was imprecise, relying on palpation of the pulse rather than listening for sounds. His method, called the auscultatory method, quickly became the clinical standard and remains the reference technique against which newer devices are validated more than a century later.
The Five Phases Explained
Clinicians and textbooks divide the sounds into five phases based on how they change in loudness, pitch, and quality as cuff pressure falls. The transitions between phases are not always knife-sharp, and some observers find them easier to distinguish than others, but the general pattern is consistent.
- Phase I: The first appearance of clear, repetitive tapping sounds. Each tap corresponds to a heartbeat where the peak arterial pressure just barely overcomes the cuff pressure, letting a small spurt of blood through. The cuff pressure at the moment you hear the first tap is recorded as systolic blood pressure. These initial sounds tend to be short and relatively crisp.
- Phase II: The tapping sounds develop a longer, swishing or murmuring quality. The artery is now open for a larger fraction of each beat, and the turbulence becomes more complex. Acoustic analysis shows that Phase II sounds have the greatest high-frequency content and the longest duration of any phase, making them visibly the most complex when viewed on a recording.2PubMed. Characterization of the Korotkoff sounds using joint time-frequency analysis This is also where the “auscultatory gap” can occur, a phenomenon discussed below.
- Phase III: The murmuring quality fades, and the sounds become loud, sharp tapping again, similar to Phase I but often louder. Sound duration drops significantly compared to Phase II.2PubMed. Characterization of the Korotkoff sounds using joint time-frequency analysis The artery is open for most of each cardiac cycle at this point, and the bursts of turbulence are shorter and punchier.
- Phase IV: The sounds suddenly become muffled and softer. The artery is nearly fully open, and the remaining turbulence is diminishing. The cuff pressure at this muffling point is sometimes used as an alternative diastolic reading, particularly in children and certain other situations.
- Phase V: Complete silence. The cuff pressure has dropped below the lowest point of arterial pressure during the cardiac cycle, so the artery stays open continuously and blood flows without turbulence. The cuff pressure at the last audible sound is the standard diastolic blood pressure reading in adults.
In practice, Phases I and V get the most clinical attention because they mark systolic and diastolic pressure. Phases II and III are mainly useful for confirming you are in the right range, and Phase IV becomes important in specific populations where Phase V is unreliable.
The Phase IV Versus Phase V Debate
For most adults, Phase V, the point of silence, is the accepted marker for diastolic blood pressure. But in some people, the sounds never fully disappear, or they fade so gradually that pinpointing exact silence is difficult. This is especially common in children. In young kids, the artery is more elastic and the vessel walls vibrate easily, so faint sounds can persist all the way down to zero cuff pressure. When that happens, Phase V is technically 0 mmHg, which is obviously not the child’s real diastolic pressure.
A large longitudinal study from the Bogalusa Heart Study found that in children under about 15, Phase IV was a more reproducible measure of diastolic pressure than Phase V, and recommended recording both values.3American Journal of Hypertension. Comparison of fourth and fifth Korotkoff diastolic blood pressures in 5 to 30 year old individuals: The Bogalusa heart study A separate analysis of over 2,000 children confirmed that Phase IV showed less variation between different observers, was more strongly linked to adult blood pressure levels, and was a better predictor of who would develop hypertension in adulthood than Phase V was.4PubMed Central. Differences between the fourth and fifth Korotkoff phases among children and adolescents In that study, 7% of children had at least one Phase V reading of 0 mmHg, underscoring why relying on silence alone is problematic in younger patients.
A study of preschool-aged children (3 to 6 years) similarly concluded that Phase IV appeared to be a better indicator than Phase V for diastolic pressure in that age group, and that the choice of phase could change how many children were classified as hypertensive.5International Journal of Cardiology. Comparison of Korotkoff phase IV and phase V for diastolic blood pressure in children aged 3 to 6 years By age 15 or so, the difference between Phase IV and Phase V narrows enough that it stops mattering clinically, and standard adult practice, using Phase V, applies from then on.
Korotkoff Sounds in Pregnancy
Pregnant women were historically another group where Phase IV was favored over Phase V for diastolic pressure, on the assumption that increased blood flow during pregnancy made Phase V unreliable. A study testing this directly found that assumption was largely a myth. Both Phase I (systolic) and Phase V (diastolic) could be detected in 98% of measurements, whereas Phase IV could only be detected in about half. In 46% of measurements, neither observer could identify Phase IV at all.6PubMed Central. The fourth sound of Korotkoff in pregnancy: a myth On top of that, observer agreement for Phase IV was poor compared to the near-perfect agreement for Phase I and Phase V. Current guidelines for blood pressure in pregnancy generally recommend using Phase V, the same as in other adults.
The Auscultatory Gap
One of the trickiest pitfalls in manual blood pressure measurement is the auscultatory gap: a silent interval that occurs in the middle of the Korotkoff sound sequence. You hear clear Phase I sounds, then the sounds disappear for a stretch of 10 to 40 mmHg as the cuff deflates, and then they reappear. If you do not inflate the cuff high enough, you might miss Phase I entirely, start listening only after the gap, and record a falsely low systolic reading. Conversely, someone deflating the cuff who hears sounds disappear during the gap might mistakenly call that silence Phase V and record a falsely high diastolic reading.
Auscultatory gaps are more common in people with high blood pressure. When systolic pressure exceeds roughly 180 mmHg, the incidence increases significantly.7Medicine in Novel Technology and Devices. A review of blood pressure measurement methods based on Korotkoff sounds Research comparing the gap phenomenon to continuous intra-arterial pressure monitoring suggests that the gap reflects real, brief dips in systolic pressure during natural blood pressure variability, not an acoustic artifact of the stethoscope or cuff. When systolic pressure momentarily drops below the falling cuff pressure, no blood pushes through and no sound is produced; when it rises again, the sounds return.8“Arterial’naya Gipertenziya” (“Arterial Hypertension”). Evaluation of blood pressure using Korotkoff’s sounds in case of auscultatory gap The practical defense against this error is simple: always inflate the cuff well above the estimated systolic pressure, and palpate the radial pulse during inflation so you know approximately when the pulse disappears.
How Technique Affects Accuracy
Even without an auscultatory gap, several aspects of technique can shift a Korotkoff-based reading by enough to change a clinical decision.
Cuff size is one of the most studied. A cuff bladder that is too narrow for the arm overestimates blood pressure, while one that is too wide underestimates it.9PubMed. The error in indirect blood pressure measurement with the incorrect size of cuff This is a common real-world problem: using a standard adult cuff on a large arm inflates the systolic reading, potentially leading to unnecessary treatment, while using too large a cuff on a thin arm might mask genuine hypertension.
Deflation speed also matters. The standard recommendation is to release cuff pressure at about 2 to 3 mmHg per heartbeat. A classic study showed that a time-based deflation rate of 3 mmHg per second works acceptably at a normal heart rate around 72 beats per minute, with a maximum error of about 2.5 mmHg. But at slower heart rates, say 40 beats per minute, the same deflation speed causes the cuff pressure to drop too far between beats, and the error climbs above 4 mmHg in both systolic and diastolic readings. Using a rate tied to the heartbeat (roughly 2 mmHg per beat) keeps the maximum error around 2 mmHg regardless of heart rate.10PubMed. The effect of cuff pressure deflation rate on accuracy in indirect measurement of blood pressure with the auscultatory method
Whether you read the sounds during inflation or deflation makes a difference as well. A study comparing both approaches found that readings taken during cuff inflation produced significantly larger systolic errors than readings during deflation, while diastolic readings during inflation were surprisingly close to true intra-arterial values.11Journal of Hypertension. CAN RECORDING KOROTKOFF SOUNDS DURING BRACHIAL CUFF INFLATION PROVIDE MORE ACCURATE NON-INVASIVE ESTIMATES OF INTRA-ARTERIAL BLOOD PRESSURE? Standard practice remains deflation-based measurement for good reason, but the finding hints at how the mechanical behavior of the artery under compression varies depending on whether pressure is rising or falling.
In people with more severe hypertension, the gap between inflation and deflation readings widens further. One study found that the systolic difference was roughly 2 mmHg in mild hypertension but nearly 10 mmHg in severe hypertension, enough to reclassify about a third of severely hypertensive patients into a lower grade when measured during inflation.12PubMed Central. Blood Pressure Difference Between Cuff Inflation and Deflation by Auscultatory Method: Impact of Hypertension Grade
How Close Are Korotkoff Sounds to True Arterial Pressure?
The gold standard for blood pressure is direct measurement through a catheter inside an artery. Korotkoff sounds track that standard well for diastolic pressure but consistently underestimate systolic pressure. A study comparing simultaneous cuff-based and intra-arterial readings in 40 patients found that diastolic values agreed almost perfectly, with an average difference of less than 1 mmHg. Systolic pressure, however, was underestimated by an average of about 14 mmHg.13PubMed. Accurate detection of Korotkoff sounds reveals large discrepancy between intra-arterial systolic pressure and simultaneous noninvasive measurement of blood pressure with brachial cuff sphygmomanometry The researchers attributed this to a mechanical delay: as cuff pressure drops just below true systolic pressure, the artery does not spring open instantly. It takes a few additional mmHg of cuff deflation before blood flow re-establishes enough turbulence to create an audible sound, so Phase I arrives a moment late.
This systolic underestimation is worth knowing about, but it does not invalidate cuff-based measurement. Virtually all the blood pressure thresholds used in clinical guidelines were developed using cuff readings, so the numbers that define “normal” or “high” already account for this offset. The problem would only arise if you tried to compare a cuff reading directly to an arterial line reading and expected them to match.
Manual Readings Versus Automated Devices
Most blood pressure monitors used at home and in many clinics are oscillometric devices. Instead of listening for Korotkoff sounds, they detect tiny pressure oscillations in the cuff itself as the artery pulses, then use an algorithm to estimate systolic and diastolic values. The auscultatory method, by contrast, relies on a trained listener identifying Phase I and Phase V directly.
A study comparing mercury manometer auscultation to automated oscillometric readings in the same patients found that the manual technique consistently produced slightly higher values, averaging about 2 mmHg higher for systolic and 1.3 mmHg higher for diastolic pressure. In roughly one in five patients, the discrepancy was large enough to be clinically relevant, and the gap was wider in patients over 65.14PubMed. Comparison of automated oscillometric versus auscultatory blood pressure measurement The concern is that someone whose pressure looks controlled on an automated device might actually be above target by manual measurement, potentially leading to undertreatment.
When both methods were compared head-to-head against intra-arterial catheter readings, the auscultatory method came out closer to the gold standard, with high correlation for both systolic and diastolic values. The oscillometric method showed poorer agreement, particularly for diastolic pressure.15PubMed Central. Comparing Intra-Arterial, Auscultatory, and Oscillometric Measurement Methods for Arterial Blood Pressure None of this means automated devices are bad; for most routine home monitoring they are perfectly adequate. But in critical care settings or when readings seem inconsistent, a skilled manual reading with a stethoscope tends to be more trustworthy.
Irregular Heart Rhythms and Other Tricky Situations
Atrial fibrillation, the most common sustained irregular heart rhythm, makes Korotkoff-based measurement particularly unreliable. When heartbeats vary wildly in timing and strength, blood pressure fluctuates from beat to beat. One heartbeat might produce a strong Korotkoff sound while the next barely registers, making it hard to pin down where Phase I truly begins or Phase V truly ends. Guidelines for people with atrial fibrillation generally recommend taking multiple readings and averaging them, since any single reading may be off by a significant margin.
Arterial stiffness, which tends to increase with age, also changes the character of Korotkoff sounds. Stiffer vessel walls transmit sound differently, and the amplitude of the sounds tends to drop as compliance decreases.16Journal of Clinical and Diagnostic Research. Using Korotkoff Sounds to Detect the Degree of Vascular Compliance in Different Age Groups In elderly patients with very stiff arteries, Phase I sounds may be faint and easy to miss, which can lead to an underestimate of systolic pressure. Inflating the cuff higher than you think necessary and listening carefully during the first few mmHg of deflation helps avoid this.
What Korotkoff Sounds May Reveal Beyond Blood Pressure
Researchers have started exploring whether the acoustic characteristics of Korotkoff sounds carry useful information beyond simple blood pressure numbers. Since the sounds are shaped by how the arterial wall vibrates, properties like frequency content and waveform shape could theoretically reflect the stiffness and health of the vessel itself. A recent study used deep learning models to classify people into age-related arterial stiffness groups based solely on features extracted from recorded Korotkoff sounds, achieving classification accuracy above 89%.17Frontiers in Cardiovascular Medicine. Assessing arterial stiffness using characteristics of Korotkoff sounds Other groups have applied machine learning to classify Korotkoff sound signals in the context of heart failure, using algorithms to denoise and segment the recordings and then detect patterns that might flag cardiac dysfunction.18PubMed Central. A Study on the Association between Korotkoff Sound Signaling and Chronic Heart Failure (CHF) Based on Computer-Assisted Diagnoses
These approaches are still in early research stages, and none have made it into routine clinical practice. But the idea is appealing: a standard blood pressure check already produces a rich acoustic signal, and if algorithms can extract vascular health information from that signal, it could turn an ordinary cuff measurement into a screening tool for arterial stiffness or heart disease without requiring any additional equipment. Whether the signal is reliable enough to add real diagnostic value over existing methods remains an open question, but the work suggests that Korotkoff sounds contain more information than Nikolai Korotkoff could have imagined when he first held a stethoscope to a compressed arm in 1905.