Walking pneumonia: how it is diagnosed and why the cough lingers

Short answer

Walking pneumonia is diagnosed clinically, not by a single test. A doctor listens, asks how long this has gone on, and usually orders a chest x-ray, because signs alone are unreliable (5). There is no quick test for the bacterium most often behind it, and most outpatients never get one (1).

If you have been coughing for three weeks and someone told you it was just a virus, you're not imagining it and you are not being dramatic. This is the single most common way walking pneumonia is missed: the illness is mild enough that you keep going to work, so nobody escalates — and the symptoms genuinely do take weeks (1). The phrase walking pneumonia exists for exactly that reason. CDC says it was coined because "with mild symptoms, people may not stay home or in bed" (1). Feeling well enough to walk around is not evidence that your lungs are clear.

Note: this article explains how doctors reach this diagnosis and is not medical advice; only a clinician who can examine you can diagnose pneumonia. Get emergency care — call 911 — for shortness of breath at rest or a struggle to breathe, chest pain or pressure, blue, gray or blotchy lips, tongue or fingertips, new confusion or being hard to wake, coughing up blood, or a home pulse-oximeter reading of 92% or less (11, 12, 14, 16). Call your doctor the same day for a fever that has not broken after 48 to 72 hours — a temperature above 102°F still there on day three certainly counts — and for a cough of around three weeks with colored phlegm, breathlessness or chest pain that comes with breathing (14, 20).

Why walking pneumonia is so easy to miss

Nothing about walking pneumonia announces itself. The bacterium usually responsible, Mycoplasma pneumoniae, has an unusually long runway: CDC says "it can take 1 to 4 weeks for symptoms to appear after someone's been exposed to the bacteria" (1). By the time you feel bad, whoever gave it to you has often forgotten being ill, so the story you bring to the appointment has no obvious starting point.

Then it builds instead of hitting. The common form is not pneumonia at all but tracheobronchitis — a chest cold with fatigue, fever, headache, sore throat "and a cough that slowly worsens," and symptoms that "can last for several weeks" (1). A cough that slowly worsens over ten days reads to everyone, including the person having it, as a cold that is dragging.

The second reason produces most of the frustration: the expectation of how long a cough should last is wrong by a factor of two or three. A systematic review paired with a patient survey found the mean duration of acute cough illness in the published literature was 17.8 days, while patients expected roughly 5 to 9 days (9). So on day 12 you feel abnormal, the reference range says normal, and the conversation ends there. That gap, not any failure of yours, is where this diagnosis gets lost.

And mycoplasma does not look like anything in particular. CDC's clinical overview for providers puts it plainly: presentations are "highly variable and non-specific" (3).

What is your doctor actually listening for?

A stethoscope is doing something real. Your doctor is listening for crackles in one area rather than everywhere, reduced air entry at a lung base, a difference between the two sides, and wheeze that would point elsewhere — while watching things you may not notice: how fast you are breathing, your heart rate, your temperature, whether you can finish a sentence.

What the exam cannot do is settle it alone, and that is why the main US guideline is written the way it is. The 2019 American Thoracic Society and Infectious Diseases Society of America guideline defines community-acquired pneumonia as "signs and symptoms of pneumonia with radiographic confirmation," and says it deliberately restricted its evidence to studies using radiographic criteria, "given the inaccuracy of clinical signs alone" (5). A guideline committee does not write that sentence about a test that works.

So: a normal-sounding chest does not rule pneumonia out, and an abnormal-sounding chest does not prove it. The exam decides whether the next step is reassurance or an image. With mycoplasma the mismatch is the whole character of the disease — people often look better than their lungs do (1).

Why would my doctor order a chest x-ray if I feel fine?

Because the x-ray answers a different question. The exam asks "does this person look unwell?" The film asks "is there consolidation in the lung tissue?" With walking pneumonia those two answers come apart more often than with any other kind.

In the US guideline the image is not an optional extra — it is part of the definition: community-acquired pneumonia means clinical features plus radiographic confirmation (5). Without the film, what you have is a chest infection of unknown depth.

One thing has moved recently. The 2025 American Thoracic Society guideline says lung ultrasound is an acceptable alternative to a chest x-ray in adults with suspected pneumonia where the expertise exists — a conditional recommendation on very low certainty evidence, not for immunocompromised patients, and published by ATS in a document that "was not approved by the IDSA" (6). If your clinic has someone who does lung ultrasound well, that may be your image. Most places will still take a film.

What an x-ray will not do is name the organism. Mycoplasma and pneumococcal pneumonia can look similar enough on a plain film that no radiologist will call it from the image. The film tells you there is pneumonia and roughly where; the organism is a separate, usually unanswered question.

And there is an afterwards worth asking about: NICE's 2025 guideline says not to routinely repeat a chest x-ray after discharge, but to consider one at six weeks for people with persisting or worsening symptoms, underlying respiratory disease, lung cancer risk factors or unexplained weight loss (15).

What a blood test can and cannot tell you

This is where expectations most often break. People arrive hoping a blood draw will say pneumonia: yes or no. It does not, and the guidelines are explicit about how little it adds in the outpatient setting.

The 2019 ATS/IDSA guideline recommends not obtaining sputum Gram stain and culture routinely in adults managed as outpatients, and recommends not obtaining blood cultures in outpatients at all — both strong recommendations (5). It also suggests not routinely testing urine for pneumococcal or Legionella antigen outside severe disease (5). The entire outpatient microbiology workup, in other words, is usually nothing.

Procalcitonin is the one people have read about. The guideline's position is that empiric antibiotic therapy should be started in adults with clinically suspected and radiographically confirmed pneumonia "regardless of initial serum procalcitonin level" — a strong recommendation (5). A normal procalcitonin does not mean you can skip treatment, and a raised one does not make the diagnosis.

A white cell count and a C-reactive protein are still worth having — they gauge how inflamed you are and give a number to repeat next week. What they cannot do is separate a bacterial chest infection from a viral one in one person, or tell you whether the thing in your chest is mycoplasma. If you left an appointment disappointed that "the bloods were normal," that is the test behaving as designed. The American Lung Association lists blood tests alongside history, listening to the lungs, chest x-ray and pulse oximetry — one input, not the verdict (21).

Why is there no quick test for walking pneumonia?

There is a test. It is simply not the kind you get at an urgent care in fifteen minutes, and for most people it would not change anything.

CDC states it directly: "There's no quick way to test for M. pneumoniae infections, unlike some other respiratory illnesses," and "many healthcare providers don't test for M. pneumoniae infection, especially for mild to moderate illness" (1). When they do test, it is usually a nose or throat swab sent to a lab (1).

Here is what each method can actually do.

PCR and other molecular tests (NAATs). These are the preferred method. CDC's laboratory page says nucleic acid amplification tests "are the preferred method of diagnostic testing" and that they "offer high sensitivity and specificity and provide timely results for treatment decisions," and can also predict antibiotic susceptibility (4). Most FDA-cleared options are multi-pathogen respiratory panels that include M. pneumoniae alongside flu, COVID-19 and RSV (4). If you are swabbed on a panel, this is probably the test you are getting.

Serology — the blood antibody test. This is the one people most often ask for, and the one with the biggest gap between what it sounds like and what it gives you. CDC says serology "lacks specificity" and "often requires multiple patient visits to collect acute and convalescent paired sera," which makes the sampling time-sensitive (4). Translated: one antibody result in week two usually cannot tell a current infection from an old one. To be meaningful it needs a second sample weeks later — by which point your cough has either settled or become a different conversation.

Culture. Reference laboratories can grow it, but CDC calls culture "time-consuming and not optimal for treatment decisions": the organism cannot be seen by light microscopy, does not cloud liquid media, and growth "can take several weeks" to confirm (4).

So the honest answer to "why didn't they test me?" is usually that the result would arrive after the illness and would not change anything. CDC does name where testing earns its place: during periods of increased community respiratory illness, especially among school-age children; in children hospitalized with pneumonia; and in anyone with pneumonia not improving on an antibiotic that cannot work against mycoplasma, such as a beta-lactam (3). That last one is the useful trigger — M. pneumoniae has no rigid cell wall, so cell-wall antibiotics cannot touch it (3). Treatment is your doctor's call; in practice a macrolide such as azithromycin is the usual outpatient prescription, and this article does not cover regimens.

Walking pneumonia vs typical pneumonia: age, onset and course

The two are not different severities of the same illness. They tend to be different organisms, in different people, with a different shape over time.

Who gets it. CDC's clinical overview says M. pneumoniae infection is most common in children, both school-age and younger, and that risk rises in crowded settings: schools, college residence halls, military training facilities, hospitals, long-term care and correctional facilities (3). Among people hospitalized with pneumonia, mycoplasma is the most common bacterial cause in children and the second most common in adults (1). Typical bacterial pneumonia, by contrast, skews older and towards people with other conditions.

How it starts. Typical pneumonia often arrives over a day or two with a high fever, shaking chills and a cough that hurts. Mycoplasma builds. The cough "slowly worsens," the fever is often modest or absent, and the headache and fatigue can be the loudest symptoms (1). In children under five the presentation changes again and may not look respiratory at all: CDC lists diarrhea, sneezing, sore throat, a stuffy or runny nose, vomiting, watery eyes and wheezing (1).

How it behaves. Most mycoplasma infections are self-limiting, serious complications are uncommon, and the infection can produce disease outside the lungs — CDC notes extrapulmonary disease can occur "without obvious respiratory disease," and tracks severity including skin disorders (2, 3). A rash alongside the cough is worth mentioning, not treating as a separate problem. "Mild" describes the average course, not your Tuesday: higher risk of serious infection goes with recovering from another respiratory illness, pre-existing lung conditions such as asthma, and a weakened immune system (3).

What else a three-week cough turns out to be

Walking pneumonia is one entry on a list, and the other entries are more common. A clinician working through a long cough is running this list in their head; knowing it helps you answer the questions they will ask.

A cough that is simply still going

Post-infectious cough is the most likely explanation. It is defined as a subacute cough lasting three to eight weeks and affects 11% to 25% of adults after a respiratory infection (10). The mechanism is mundane: the infection leaves the airway inflamed and twitchy, mucus up and clearance down. It is "time limited and self resolving," and trials of inhaled steroids, bronchodilators and oral agents have not shown benefit (10).

The useful number is eight weeks: beyond that a cough is chronic and needs assessment such as lung function testing (10).

Whooping cough

Pertussis is the one most likely to be mistaken for walking pneumonia in an adult, because the early phase is indistinguishable from either. CDC says "whooping cough appears similar to a common cold early on," with a runny nose, a low-grade fever and a mild occasional cough lasting one to two weeks (8). Then coughing fits begin, usually last one to six weeks, and can last up to ten (8).

CDC also says the quiet part out loud: "healthcare providers often don't suspect or diagnose it until more severe symptoms appear" (8). What pushes it up the list: coughing fits rather than a constant cough, vomiting after a fit, and a whoop on breathing in (8, 10). If there is a newborn in your household, say so — babies with pertussis often do not cough at all and may stop breathing instead (8).

Flu, COVID-19 and RSV

All three can look like this, and none can be sorted out by symptoms. CDC states that "you cannot tell the difference between flu and COVID-19 by symptoms alone" (12), and that RSV symptoms "are not specific, which makes it difficult to distinguish RSV from the common cold or other respiratory viruses" (13).

Onset is the one clue worth offering. Flu symptoms "usually come on suddenly," against the gradual onset of a cold (11); mycoplasma is the opposite end of that scale. "I know the hour it started" points towards flu; "it crept up over ten days" points away. All three can also leave a cough behind after the virus clears, which is how a flu in week one becomes a post-infectious cough in week four. See when to get a flu shot and long COVID symptoms.

Asthma that shows up as a cough

Asthma does not always wheeze. In a long cough it is one of the conditions that has to be excluded before post-infectious cough can be called, along with COPD, reflux and ACE-inhibitor use (10). The pattern to report: cough that is worse at night, worse with cold air or exercise, or that started after a chest infection and never quite stopped. Pre-existing asthma also raises the risk of a more serious mycoplasma infection (3), so it belongs in the history either way.

What a home pulse oximeter can and cannot settle

A fingertip oximeter is the only home device that reads on the thing that actually matters in pneumonia, so it deserves a careful paragraph.

What it can do: give you a number that is hard to argue with. Home-monitoring guidance written for exactly this situation tells people to call 999 — the UK emergency number — for a reading of 92% or less, and to contact a clinician at 93% or 94% (16). That protocol was written for COVID-19, but the thresholds are the ones home oximeter users are given; in the US the equivalent is 911 or an emergency department.

What it cannot do is be trusted to read low when it should. A retrospective analysis found occult hypoxemia — true arterial saturation below 90% while the oximeter showed 92% or higher — in 11.7% of Black patients against 3.6% of White patients, roughly three times as often (17, 18). The Anesthesia Patient Safety Foundation notes that analysis was retrospective and without objective skin-tone measurement, and gives the instruction worth keeping: do not base a decision on a single SpO2 value, and remember that perfusion, anemia, motion and the device itself all affect it (17).

So the rule is asymmetric. A low reading is a reason to act. A normal reading is not a reason to stay home if you feel wrong. NICE's 2025 guideline treats saturation below 90% on room air as clinical instability in people being treated for pneumonia — a hospital-side threshold, not a reassurance (15). Breathlessness at rest, chest pain with breathing, blue or gray lips and new confusion are emergencies whatever the screen says (11, 12, 14).

When walking pneumonia circulates: seasons, cycles and the 2024 surge

Most respiratory advice runs on a fall-to-winter calendar: RSV "usually peaks in December and January" (13), and flu shots are timed to match. Mycoplasma does not follow that calendar, and that is part of why it surprises people.

CDC's surveillance page says peaks typically occur every three to seven years and that infections "may be more common in summer and early fall" (2); CDC's clinical overview puts the cycle at "about every 4 years" (3). Two CDC pages updated the same day, two different numbers — which tells you how soft the estimate is. There is no national reporting system for M. pneumoniae, so true case numbers are unknown (3).

The recent wave is well documented: CDC identified an increase through 2024, rising through the summer and "peaking for the 2–4-year-old and 5–17-year-old age groups in August 2024," notable because these infections had been thought of as a school-age problem. Children under five in 2024 did not have more severe illness than in 2018–2023 (2). As of CDC's surveillance update of May 18, 2026, infections have been decreasing since the beginning of 2025 but "remain elevated in some regions" (2). CDC also monitors macrolide resistance — about 28% globally, under 10% in the US (2) — one reason a clinician may want a result when someone is not improving.

The practical consequence is small but real: a cough in April is not evidence against walking pneumonia, and a negative flu test in November is not evidence against it either.

Living with a cough that will not go: the questions people ask next

How long does walking pneumonia last, and when will I feel normal again?

Longer than you want. CDC says symptoms of M. pneumoniae infection "can last for several weeks," and recovery is gradual (1). The cough leaves last: the mean duration of an acute cough illness in the literature is 17.8 days, and a post-infectious cough can run three to eight weeks (9, 10). Hold onto eight weeks — past that, a cough counts as chronic and needs reassessment rather than more patience (10).

Is walking pneumonia contagious, and should I stay home?

Yes, it spreads — by airborne droplets through close person-to-person contact, which is why outbreaks cluster in schools, dorms, military training facilities and long-term care (3). The awkward part is timing: with an incubation period of one to four weeks, you were probably spreading it before you thought of yourself as ill (1). Your workplace rules and your clinician decide when you go back. If you live with a newborn, someone pregnant or anyone immunocompromised, raise that at the appointment.

Can walking pneumonia go away on its own?

Often, yes. CDC says most M. pneumoniae infections are self-limiting, while also noting that providers routinely treat mycoplasma pneumonia with antibiotics (3). Both are true: self-limiting describes the average course, treatment is a judgement about you. What you should not assume is that because one antibiotic did nothing, nothing will — some common antibiotics cannot work against this organism, which has no cell wall for them to attack (3). "It didn't help" is information, not a dead end.

I feel fine apart from the cough. Can I train?

That is a decision for you and your clinician, and the honest answer depends on whether this is a chest infection or only its echo. Fatigue with walking pneumonia is often out of proportion to how your chest feels, and one of the risk factors CDC lists for a more serious infection is already recovering from another respiratory illness (3). We have a separate guide on working out while sick that covers how to think about this properly; it is a different question from the diagnostic one this article is about.

Why is the cough worse at night?

A cough that peaks when you lie down is a common complaint during recovery from walking pneumonia, and it is also the pattern that should make a clinician think about asthma, reflux or post-nasal drip rather than leftover infection — exactly the conditions that must be excluded before a lingering cough is called post-infectious (10). Note whether it is dry or productive, whether it wakes you, and whether anything makes it predictable; that is more useful at an appointment than "it's bad at night."

This is my third chest infection this year. Is something wrong with me?

A fair question, and worth putting in front of a doctor if each episode has been long. Walking pneumonia does not explain a pattern of repeat infections; recurrent pneumonia is one of the red flags that justifies imaging, alongside coughing up blood, weight loss or an extended smoking history (10), and NICE lists persisting symptoms and underlying respiratory disease as reasons to consider a follow-up chest x-ray at six weeks (15). Our piece on why you keep getting sick covers the broader pattern.

Should my kids be tested if I have it?

That is their pediatrician's decision, but the context is useful. Mycoplasma is most common in children, and CDC suggests considering testing for it alongside other respiratory pathogens during periods of increased community respiratory illness, especially among school-age children, and in children hospitalized with pneumonia (3, 4). In children under five the illness may not look like a chest infection at all — diarrhea, vomiting, sneezing and watery eyes are on CDC's list (1). Keeping their respiratory vaccinations current is a separate conversation; see our guide to the flu shot for kids.

How to bring this up with your doctor

Before you blame yourself for making a fuss about a cough: the thing that most often decides this diagnosis is not a clever test, it is whether the timeline in your head makes it onto the record. Here is how to make sure it does.

Bring two numbers and one date. The date the first symptom started — not the date it got bad. How many weeks it has now been. And, if you track anything at all, what your resting heart rate has been doing. "Three weeks tomorrow, started with a sore throat on the 24th, and my resting heart rate has been about six beats above my usual since then" is a different consultation from "I've had this cough for ages."

Say the sentence that moves things. Not "I think I have pneumonia," but: "This cough is in its third week and it is not improving. Do you think a chest x-ray is warranted?" The 2019 US guideline defines pneumonia as clinical signs plus radiographic confirmation, so the image is the question the diagnosis turns on (5). Where a trained clinician offers lung ultrasound, the 2025 ATS guideline considers that an acceptable alternative (6).

Name the specific test, not "tests." For the organism: "Would a respiratory PCR panel including Mycoplasma pneumoniae change anything here?" Molecular testing is the preferred method and usually runs as a multi-pathogen panel (4). Be ready for a reasonable no — in mild outpatient illness most people are not tested (1). "A blood test to check for pneumonia" is the request most likely to go nowhere, because outpatient blood work does not settle it (5).

Flag the three triggers that change the answer. Say them out loud if they apply: I have already had an antibiotic and it did nothing (mycoplasma is not touched by cell-wall antibiotics, and non-response is a recognised reason to test); I have asthma or another lung condition, or my immune system is suppressed (higher risk of serious infection); there is a baby, a pregnant person or someone immunocompromised at home (3).

If you are brushed off. Ask for the reasoning to be written down: "Can we note in my record that we decided against imaging today, and agree what would change that?" Then set the follow-up yourself. NHS guidance says to ask for an urgent appointment for a cough of around three weeks with colored phlegm, chest pain on breathing, or breathlessness (14); eight weeks is the point at which a cough is formally chronic and needs assessment, not more waiting (10). Both are dates you can book in advance.

How Welltory helps — and what it cannot do

The limits first, and here they are not a formality. Welltory cannot find pneumonia. It is a general wellness product, not a medical device: it does not diagnose, detect, predict, monitor, prevent, treat or mitigate any condition, including any chest infection. It does not measure blood oxygen and is no substitute for a pulse oximeter or a chest x-ray, and nothing in it should postpone an appointment. Welltory's help center also notes measurements may be less accurate for people under 18 and for pregnant people (19).

What it can do is answer the question your doctor will actually ask: how many weeks has this been going on, and is it getting better or worse? Memory is bad at this. A dated log and a flat daily record of resting heart rate answer it more precisely than recall — and that gap is often the difference between "a cough" and "a cough in its third week that is not improving."

You note what happened; the app keeps the rest. Tap the plus at the top of the feed to add a tag: "day 1 cough," "fever 101," "no better," "started antibiotic." When the Today screen marks a stress or rest stretch it asks "What happened?" — tap a suggested tag, type or talk, and correct the tags before saving.

Then look at the days around them. Resting heart rate, heart rate variability, stress minutes, Battery and sleep are recorded the same way every day, so sick days sit next to your own ordinary days rather than a textbook range (Battery and stress minutes need iOS with an Apple Watch or Oura; detailed sleep analysis needs an Apple Watch). If you do not know your own baseline, start with what counts as a normal resting heart rate.

Over time, My Patterns looks for what your tagged days have in common (iOS with an Apple Watch or Oura; patterns start appearing at around seven tagged events in a month, so a single illness will not produce one). Anything it finds is an association worth discussing, never a cause and never a diagnosis.

A record you can hand over. The Journal (Premium) keeps your measurements, tags and notes, and from the web app you can export a CSV (Dashboard → choose a chart → Export). Walking in with a dated line is not evidence of pneumonia. It is evidence of duration and direction, which is exactly what the exam cannot supply.

One thing we are not showing you. We have not published a dataset on how resting heart rate behaves across tagged sick days, so there is no chart of our own here — every number above comes from the published sources listed below. Better to say so than draw a line we cannot stand behind.

The same red flags apply here, in the same type size: shortness of breath at rest, chest pain or pressure, blue or gray lips or fingertips, new confusion, coughing up blood, or a pulse-oximeter reading of 92% or less means emergency care now — whatever your app says (11, 14, 16).

How we made it

Made with AI tools, then edited and fact-checked by the Welltory team. See our Editorial & AI policy.

Data analysis by Jane Smorodnikova, co-founder of Welltory and the person who built the methodology behind how we read physiological data.

Written by Tatsiana Yashyna.

Discounts for blog readers: up to 36% off

See what affects your energy, stress, sleep, and daily state with Welltory

Explore your data