Korotkoff sounds are the thumping, swooshing, and tapping noises you hear through a stethoscope while deflating a blood pressure cuff, and they have been the foundation of manual blood pressure measurement for well over a century. They progress through five distinct phases as cuff pressure drops, with Phase I marking the systolic reading and Phase V marking the diastolic reading in most patients. Understanding what each phase sounds like and why it occurs gives you a much clearer picture of what is actually happening during a blood pressure check, and it helps explain why certain situations, like pregnancy or advanced age, can make the measurement tricky.
How Korotkoff Sounds Are Generated
When you inflate a blood pressure cuff above the systolic pressure in the brachial artery, blood flow is completely blocked. As you slowly release air from the cuff, the external pressure drops below peak arterial pressure, and a small jet of blood starts squeezing through the partially compressed artery with each heartbeat. That turbulent, pulsatile flow is what creates sound. The artery wall itself also vibrates as it snaps open and then collapses again during each cardiac cycle. These combined vibrations travel through tissue and can be picked up with a stethoscope placed just below the cuff.
As cuff pressure continues to fall, the degree of arterial compression changes, and so does the character of the turbulence. The sounds shift in pitch, loudness, and quality until the artery is no longer compressed at all, at which point laminar flow resumes and the sounds disappear. That progression from first audible tap to silence is what clinicians divide into five phases.
Arterial wall stiffness plays a role in how well these sounds transmit. In older adults, the vessel walls tend to be thicker and less elastic, which can reduce the amplitude of the transmitted sounds and make the phases harder to distinguish.
Phase I Through Phase V
Phase I
Phase I is the first clear, repetitive tapping sound you hear as you deflate the cuff. It corresponds to the point where cuff pressure drops just below peak systolic pressure, allowing small bursts of blood to push through the compressed artery. The reading on the gauge at the moment you first hear this tap is recorded as the systolic blood pressure. These initial sounds are relatively sharp and distinct, which is why Phase I is generally the easiest to identify. In clinical practice, if someone says “120 over 80,” that 120 came from Phase I.
Phase II
As the cuff deflates further, the sounds develop a softer, swishing or murmuring quality. Phase II sounds are longer in duration and lower in intensity than the crisp taps of Phase I. The turbulence is now happening over a wider portion of the cardiac cycle because the artery stays open a bit longer with each beat. For most routine blood pressure checks, Phase II is simply something you listen through on your way to the diastolic endpoint. But it takes on clinical importance when it disappears prematurely, a phenomenon called the auscultatory gap, which is discussed below.
Phase III
Phase III sounds are louder and crisper again, resembling the tapping of Phase I but often with more intensity. The artery is now open for a larger portion of each heartbeat, producing stronger vibrations. Some training materials describe Phase III as a return to the thumping character heard at the start, but with greater volume. In practice, many clinicians find the boundary between Phase II and Phase III hard to pinpoint, and for standard blood pressure recording it does not need to be identified precisely.
Phase IV
In Phase IV the sounds abruptly become muffled, softer, and take on a blowing quality. The artery is now barely compressed, so the turbulence that generates sharp sounds is fading. Phase IV is sometimes described as a sudden “dulling” of what had been loud, clear tapping. For most adults, Phase IV is simply a brief transition before the sounds vanish entirely. But in certain patient populations, Phase IV becomes the clinically relevant diastolic endpoint, because the sounds never fully disappear.
Phase V
Phase V is silence. It marks the point where cuff pressure has fallen below the diastolic pressure in the artery, the vessel stays open throughout the entire cardiac cycle, laminar flow is restored, and no more turbulent sound is generated. The gauge reading at the last audible sound is recorded as the diastolic blood pressure. Research confirms that the disappearance of Korotkoff sounds closely approximates true intra-arterial diastolic pressure, which is why Phase V is the standard diastolic endpoint for adults.
The Auscultatory Gap
Sometimes, after the initial Phase I taps appear, the sounds vanish for a stretch of cuff deflation and then reappear at a lower pressure. This silent window is the auscultatory gap, and it can lead to serious measurement errors if you are not aware of it. If you start listening only after the gap has begun, you might mistake the reappearance of sound for Phase I and record a falsely low systolic reading. Conversely, if you stop deflating when the gap opens, you could record a falsely high diastolic number.
A study of hypertensive patients found auscultatory gaps in about one in five, with the gaps linked to older age, female sex, greater arterial stiffness, and atherosclerotic plaque in the carotid arteries. Even after adjusting for age, arterial stiffness and the presence of plaque remained independently associated with gaps. The prevalence of atherosclerotic plaques was more than doubled in patients who had gaps compared with those who did not.
The practical takeaway is straightforward: always inflate the cuff well above the expected systolic pressure, typically by palpating the pulse and inflating at least 20 to 30 mmHg above the point where the pulse disappears. That way, you catch Phase I above the gap and do not miss the true systolic reading.
The Diastolic Dilemma in Pregnancy, Children, and High-Output States
In most adults, the sounds reliably vanish at Phase V, giving you a clean diastolic number. But in some situations, Korotkoff sounds persist all the way down to zero cuff pressure, meaning Phase V never arrives. This happens when increased blood flow or altered vascular dynamics keep turbulence going even in a fully open artery. The classic examples are pregnancy, severe anemia, hyperthyroidism, aortic valve insufficiency, and sometimes children, where vascular elasticity and cardiac output differ from a typical adult.
When Phase V is absent, clinicians use Phase IV, the muffling point, as the diastolic reading instead. A review of measurement methods describes this as the “diastolic measurement dilemma” inherent to auscultatory techniques: in patients with high-cardiac-output states, diastolic blood reflux intensifies the turbulence enough to overpower the normal disappearance of sound, so the muffling point becomes the only available endpoint.
Pregnancy has been a particular battleground for this question. For years, some guidelines recommended Phase IV as the standard diastolic endpoint in pregnant women because of reports that Phase V occasionally reached zero. However, research comparing the two phases against direct intra-arterial measurements found that Phase V actually tracks true diastolic pressure more closely in pregnancy.
One study reported that the average difference between Phase IV and Phase V fluctuated around 6 mmHg throughout pregnancy, with Phase V showing a stronger association with clinical outcomes related to hypertension, including proteinuria, fetal growth restriction, and elevated uric acid. A separate analysis similarly concluded that available evidence supports Phase V as the preferred endpoint in pregnant women. The current consensus in most guidelines is to use Phase V in pregnancy unless the sounds genuinely persist to zero, in which case Phase IV is the fallback.
Why Cuff Deflation Speed Matters
The standard recommendation is to deflate the cuff at roughly 2 to 3 mmHg per second or per heartbeat. This sounds like a minor technical detail, but it directly affects accuracy. If you let air out too fast, you can blow right past the true Phase I or Phase V point and record a number that is several mmHg off.
Modeling work on the relationship between deflation rate and measurement error showed that at a standard rate of 3 mmHg per second, the maximum error in both systolic and diastolic readings can reach about 2.5 mmHg at a normal heart rate. At a slow heart rate of 40 beats per minute, that error can climb above 4 mmHg because fewer heartbeats occur during each mmHg of pressure drop, giving you fewer chances to hear the transition points. A rate based on heartbeats rather than seconds, around 2 mmHg per beat, keeps the error at roughly 2 mmHg regardless of heart rate.
Research comparing fast and slow deflation directly found that for manual auscultatory measurement, accurate readings were achievable only at the recommended slow rate. Faster deflation also increased variability between repeated measurements on the same person, meaning not only were the readings less accurate, they were less consistent.
The degree of hypertension itself amplifies the problem. A study examining inflation versus deflation readings found that the discrepancy between the two grew progressively with the severity of hypertension, jumping from under 2 mmHg in mild cases to nearly 10 mmHg in severe hypertension for systolic readings.
Pseudohypertension and Stiff Arteries
In some older adults, the arteries become so calcified and rigid that the blood pressure cuff cannot fully compress them. The result is that the cuff registers a much higher pressure than the actual pressure inside the artery, a phenomenon called pseudohypertension. These patients appear to have dangerously high blood pressure on a standard reading, but their true intra-arterial pressure is considerably lower.
A landmark study described a bedside test called Osler’s maneuver, where you inflate the cuff above the systolic reading and then try to palpate the radial or brachial artery downstream. In a normal artery, no pulse should be felt once the artery is fully compressed. In a heavily calcified artery, the vessel wall remains palpable even when blood flow has stopped because the wall itself is too rigid to collapse. Patients who were Osler-positive had cuff readings that overestimated their true blood pressure by 10 to 54 mmHg. The study found that pseudohypertension is common in the elderly and becomes more pronounced as arterial compliance decreases.
This matters for Korotkoff sound interpretation because the entire auscultatory method assumes the cuff can fully occlude the artery. When it cannot, every phase is shifted upward, and the Phase I and Phase V readings both overestimate the true values. Clinicians who suspect pseudohypertension may need to confirm with direct arterial measurement.
Observer Bias and Terminal Digit Preference
Even when the sounds are textbook-perfect, the human listener introduces error. One well-documented problem is terminal digit preference, where the person recording the reading unconsciously rounds to a favorite number. A study at a specialty hypertension clinic found that zero was the terminal digit for systolic readings 40% of the time among nursing staff and 31% of the time among physicians. For diastolic readings, the pattern persisted at 23% for nurses and 36% for physicians. If readings were truly random, zero should appear as the final digit only about 10% of the time.
This kind of rounding does not sound dramatic, but when clinical decisions hinge on whether a patient’s blood pressure is, say, 138 or 142, rounding to 140 in either direction can push someone across a treatment threshold. The study’s authors noted that even in a specialist hypertension unit staffed by people who should know better, digit preference was measurable. Regular auditing of recorded readings and feedback to staff are the main countermeasures, though automated devices sidestep the problem entirely by generating a number without human judgment.
How Oscillometric Devices Handle Things Differently
Most home and many clinical blood pressure monitors do not use Korotkoff sounds at all. They use the oscillometric method, which detects tiny fluctuations in cuff pressure caused by arterial pulsations. Instead of listening for sound onset and disappearance, the device’s algorithm identifies the point of maximum oscillation and then uses proprietary formulas to calculate systolic and diastolic values from the shape of the oscillation waveform.
The auscultatory technique remains the reference standard against which these oscillometric devices are validated. But that validation process has gaps. A review of non-invasive measurement devices pointed out that many commercially available oscillometric monitors have never been formally validated, and those that have use undisclosed algorithms to translate oscillation patterns into systolic and diastolic numbers. The same review noted that even the exact origin of Korotkoff sounds is still debated in the research literature, and that comparing auscultatory readings with intra-arterial measurement reveals its own limitations.
In practice, this means an oscillometric monitor and a skilled clinician with a stethoscope can give you slightly different numbers for the same arm at the same moment, and neither is guaranteed to perfectly match what is happening inside the artery. For most people, the differences are small enough to be clinically irrelevant, but in edge cases like severe arrhythmia, very stiff arteries, or obesity with a poorly fitting cuff, the discrepancies can be meaningful.
Arm Position and Sound Loudness
A practical trick that is supported by research but rarely mentioned in standard guidelines involves arm elevation before measurement. A study found that raising the arm overhead for 30 seconds before inflating the cuff made the first Korotkoff sound nearly twice as loud on average compared with keeping the arm in the standard position. The second and third Korotkoff sounds were also louder.
The likely explanation is that elevating the arm drains venous blood out of the forearm, reducing the tissue volume and bringing the artery closer to the skin surface and the stethoscope. This can be genuinely useful in patients who have faint Korotkoff sounds, such as obese individuals, those with low blood pressure, or elderly patients with stiff arteries that transmit sound poorly. The catch is that the arm must be returned to heart level before the reading is taken, because the hydrostatic pressure difference from elevation would otherwise artificially lower the recorded number. The technique is about improving audibility, not changing the hemodynamics during measurement.
Korotkoff Sounds as a Window Into Arterial Health
Beyond their role in blood pressure measurement, the acoustic characteristics of Korotkoff sounds may carry information about the condition of your arteries. With increasing age, the vascular wall becomes thicker and less compliant, which reduces the amplitude of the transmitted sounds. Researchers have been exploring whether the frequency content, timing, and shape of Korotkoff sound waveforms can serve as a non-invasive marker of arterial stiffness, a known risk factor for cardiovascular disease.
A recent study using deep learning models to classify Korotkoff sound features achieved classification accuracies above 89% when sorting participants into age-stratified groups, outperforming traditional feature-based analysis. When the same approach was tested against a direct measure of arterial stiffness, the association was statistically significant but modest. The technology is still in early stages, but the idea is appealing: every blood pressure check already generates Korotkoff sounds, so extracting additional vascular health data from them would cost nothing extra in terms of equipment or patient discomfort.
Characterizing Korotkoff sounds with joint time-frequency analysis has shown that the phases differ measurably in their time, frequency, and amplitude profiles. This kind of signal processing research is laying the groundwork for smarter stethoscopes or cuff-based devices that could eventually flag abnormal arterial stiffness during a routine blood pressure measurement, turning an everyday clinical task into a screening opportunity.