Neutrophils High, Lymphocytes Low, CRP Raised

A neutrophil-dominant rise in white cells with a suppressed lymphocyte count and a markedly raised CRP leans toward a bacterial process rather than a viral one. It leans that way; it does not settle the question. Several common situations that have nothing to do with infection produce most of the same picture, which is why this combination has to be weighed against your symptoms.

The pattern on your report

  • White cell count High · moderate Supporting
  • Neutrophils High · moderate Key
  • Lymphocytes Low · mild Key
  • CRP High · marked Key

Printed as: CRP in mg/Lor mg/dLLymphocytes in x10^9/Lor x10^3/uL— Same two notations, same number, and the same split between an absolute count and a percentage.Neutrophils in x10^9/Lor x10^3/uL— The two notations carry the same number. Reports also give a percentage of total white cells, which moves independently of the absolute count. The absolute count is the one read here.White cell count in x10^9/Lor x10^3/uL— The two notations are numerically identical.

Why the numbers look like this

A white cell count is a sum, and the sum hides the split. Neutrophils and lymphocytes are recruited by different signals, which is why they are counted separately: which of the two rose, and what the other one did at the same time, carries more information than the total does. CRP is made by the liver in response to those same inflammatory signals, so it belongs beside the differential.

Almost all of these results are explained by something near the top of the list below; the serious causes are at the bottom because that is how often they occur.

Not being flagged is not the same as normal

A total white cell count can sit inside the reference range while the split beneath it is markedly skewed. A high neutrophil count and a low lymphocyte count add up to something unremarkable. So the differential is worth reading even when the headline number carries no flag beside it.

What else on the report can hide this

Corticosteroids are the classic mimic for two of these three findings, and the classic trap for the third. They drive neutrophils into the circulation and suppress lymphocytes, so the white cell picture looks like this one. But they also damp the interleukin-6 signal that drives CRP production, and a falling CRP is often how a steroid's effect is tracked. So a recent steroid explains a neutrophil rise with low lymphocytes; it does not explain a markedly raised CRP sitting beside them. Steroid plus a high CRP is a reason to look harder for an infection, not to relax.

A lymphocyte count that is still low weeks after the illness has settled is a separate question from the acute picture. Persistent lymphopenia has causes of its own, HIV among them, and it needs a test rather than more waiting.

What usually causes it

Listed from most to least common — not from most to least serious.

  1. Very common

    Bacterial infection

    Localizing symptoms and fever. A CRP above roughly 100 mg/L and a procalcitonin above about 0.5 ng/mL both shift the odds toward bacterial, though neither settles it alone.

  2. Common

    Viral infection

    Usually lymphocyte-predominant with a normal or low total white count, but influenza and COVID-19 typically drop the lymphocytes and raise the CRP, sometimes with a neutrophil rise as well. A CRP that stays modest in someone who is clearly unwell argues viral.

  3. Common

    Corticosteroid use

    Steroids push neutrophils off the vessel walls into the circulation and drop lymphocytes at the same time, so they reproduce the white cell half of this pattern. They lower CRP rather than raise it, so they do not account for a markedly raised one. Ask about any recent oral steroid, injection, or taper.

  4. Common

    Tissue injury or sterile inflammation

    The first days after surgery, a heart attack, a pulmonary embolism, pancreatitis, a gout flare or alcoholic hepatitis. CRP can pass 100 mg/L with no infection anywhere; the history of the preceding few days is what identifies it.

  5. Common

    Physiological stress or demargination

    A seizure, strenuous exercise, acute pain or fright, or the first hours after an injury. Rises quickly, settles within hours to a day or two, and on its own does not push the CRP far.

  6. Common

    Smoking

    Produces a persistent mild neutrophilia, visible on older panels going back years. But it raises the lymphocyte count rather than lowering it, and the CRP rise it causes is slight, so it accounts for only part of this picture.

  7. Uncommon

    Autoimmune or inflammatory flare

    Raised CRP without a source of infection, on a background of known inflammatory disease.

  8. Uncommon

    Malignancy

    A solid tumor or lymphoma can drive a persistent neutrophilia with a raised CRP and no source of infection. Weight loss, night sweats, or a count that never settles are what raise it.

  9. Rare

    Hematologic disorder

    Considered when the white count is very high, does not settle, or the differential shows immature forms.

Why the answer changes what you take

Antibiotics act on machinery that only bacteria have: a cell wall, bacterial ribosomes, bacterial DNA enzymes. Viruses have none of it. An antibiotic taken for a viral infection will not shorten it by a single hour, but it still carries every one of its side effects and still drives resistance. That asymmetry is why this distinction matters.

If it is bacterial

What treats it

Antibiotics. Which one depends on the likely organism, the site of the infection, local resistance patterns, your allergies, your kidney function and whether you are pregnant. That is a prescriber's decision, and no lab result can make it.

What does not

Antivirals do nothing here. Whether an antibiotic is needed at all depends on the site and the severity, because some bacterial infections settle on their own. That is part of why a delayed prescription is sometimes the right call.

Treatment is often started empirically before cultures return, then narrowed once the organism is known. When someone is unwell, that is standard practice rather than overtreatment.

If it is viral

What treats it

Supportive care: fluids, rest, and fever or pain relief. Specific antivirals do exist for a small number of viruses: influenza, COVID-19 and the herpes family. Those are time-critical.

What does not

Antibiotics. No benefit whatsoever against a virus, however severe the illness and however many weeks it drags on.

A secondary bacterial infection can develop on top of a viral one. That is why a clear worsening after you had started to improve is worth a re-check rather than a wait.

Taking the wrong one is not neutral

  • Side effects are common: most often nausea and diarrhea, sometimes allergic reactions that can be severe.
  • Clostridioides difficile colitis. Antibiotics wipe out the protective gut bacteria and let C. difficile overgrow. It can appear weeks after the course finished, and it can be serious.
  • Resistance is personal, not only societal. Your own gut carries resistant organisms for months afterwards, which matters if you later need an antibiotic that works.
  • Some classes carry risks of their own. Both the FDA and the EMA have restricted fluoroquinolones over tendon rupture and nerve damage.
  • An antibiotic that was never going to help can delay finding the actual cause.

What not to do

  • Never take leftover antibiotics from a previous illness. Likely the wrong drug, in an incomplete amount, and possibly expired.
  • Never take a prescription issued to someone else.
  • Never give aspirin to a child or teenager with a viral illness. The association with Reye syndrome is why pediatric fever is managed with other options.
  • Never shorten or extend a course on your own judgment. Course lengths are chosen deliberately, and current guidance often uses shorter courses than people expect.

Where timing matters

  • Influenza Antivirals work best when started within about 48 hours of symptom onset. If flu is suspected, being seen early changes what is possible.
  • COVID-19 Oral antivirals are given within about 5 days of symptom onset, and only for people at higher risk of severe disease. The window closes quickly here too.
  • Suspected sepsis or meningitis Hours matter. This is an emergency, not a discussion about which antibiotic.

Worth asking aboutDelayed prescribing: a prescription you hold and fill only if you are not improving by an agreed point. It is an established option, and worth raising when the picture is genuinely uncertain.

What is usually checked next

  • Blood film / manual differential Identifies atypical lymphocytes (viral) or blast cells (needs escalation).
  • Cultures and site-directed imaging Locates the source when one is suspected.
  • Repeat full blood count in 24-72 hours Separates a transient stress response from an ongoing process.
  • Procalcitonin Sometimes used to add weight to the bacterial-versus-viral question when it is raised. It adds modest weight, it does not decide, and guidelines differ on how much to lean on it.
  • HIV test, if the lymphocyte count stays low once the illness has settled A lymphocyte count that has not recovered after the acute illness is its own question, and HIV is part of the answer to it.

When to seek care sooner

  • Emergency A rash that does not fade when pressed under a glass, with neck stiffness, severe headache or dislike of bright light
  • Emergency Fever with a fast heart rate, low blood pressure, or new confusion
  • Emergency A white cell count that is very high or very low, alongside fever
  • Same day Blast cells reported on the blood film
  • Soon Raised inflammatory markers persisting with no identified source

Questions worth bringing to your appointment

  1. Does this pattern point to a bacterial cause strongly enough to justify antibiotics, or is it worth waiting and repeating the count?
  2. If it is uncertain, is a delayed prescription an option rather than starting antibiotics today?
  3. Could any medication I am taking , a steroid in particular, be producing this picture?
  4. Would a procalcitonin help decide?
  5. When should this be rechecked to confirm it is settling?

More from this panel

How to read a full blood count →

Related patterns