Read this before the chart

This blood test results chart gives typical adult ranges printed by NHS laboratories for the lines a routine set of results most often carries: the full blood count, glucose and HbA1c, kidney and liver tests, TSH and ferritin. Use it as a reading aid, not a verdict: the range that applies to your value is the one printed beside it on your own results.

For most tests there is no single national range. Each laboratory sets or checks its own, because the figure depends on the analyser, the reagents, the method and the people it serves. Two NHS trusts can print different limits for the same test and both be right: one laboratory in Leeds gives adults a single ferritin range of 30 to 337 µg/L, while one in Gloucestershire prints 24 to 340 for men and 11 to 310 for women.

The UK has worked to narrow these gaps. The Pathology Harmony project, supported by the Department of Health in England, has agreed common ranges for many routine chemistry tests, such as sodium at 133 to 146 mmol/L and potassium at 3.5 to 5.3 mmol/L in adults. Laboratories are under no legal obligation to adopt them, and blood count and hormone ranges still vary more.

How a laboratory builds a range

To set a range, a laboratory tests healthy people, at least 120 of them for each group according to the usual method, and keeps the band that contains 95% of their results, 19 in every 20. The one healthy person in twenty who lands outside that band is not ill; they sit at the edges.

That has a consequence nobody warns you about. On a panel of 20 tests, a 2025 review in the Journal of the American Board of Family Medicine calculates that roughly two healthy people in three will have at least one result marked high or low. British figures point the same way: at one Birmingham hospital, 30% of the liver test requests sent by GP practices in 2016 came back with at least one abnormal result.

The same review notes that minor variations labelled high or low cause patients worry it calls unwarranted. The flag is doing its job: it says the value sits at the edge of what the laboratory saw in healthy people, and nothing more.

Full blood count

Adult values from one NHS trust. UK results give haemoglobin in g/L rather than g/dL, so 135 g/L is the same as 13.5 g/dL on an American report. White cells and platelets are given as ×10⁹/L.

Haemoglobin is where sex matters most. British gastroenterology guidance, following the World Health Organization, defines anaemia as a haemoglobin below 130 g/L in men and below 120 g/L in non-pregnant women, but asks clinicians to use the lower limit of the laboratory that did the test. NHS trusts print a floor of 115 g/L for women in Gloucestershire and 120 g/L in Leeds, which is exactly the kind of gap that makes your own results the reference.

Read the count as a group rather than line by line: haemoglobin with MCV, which describes red cell size, and with ferritin; white cells with their breakdown into types; platelets with the rest. A single line just outside its band is a different situation from several lines moving together, and that difference is a fair question to put to your GP.

Haemoglobin

men

130 – 180 g/L

Haemoglobin

women

115 – 165 g/L

MCV

red cell size

80 – 100 fL

White cells

total count

3.6 – 11.0 ×10⁹/L

Platelets

140 – 400 ×10⁹/L

Reference bands shown as a guide. The range printed beside your own result is the one that applies.

Glucose, kidney, liver, thyroid and iron

UK results give glucose and cholesterol in mmol/L, creatinine in µmol/L and HbA1c in mmol/mol. Where a target or a risk threshold replaces a range, as it does for cholesterol and LDL, the figure comes from NICE guidance and your own cardiovascular risk, not from the laboratory.

Creatinine is made by muscle as well as cleared by the kidneys, so the eGFR worked out from it is a calculation, not a measurement. NICE asks UK laboratories to use the CKD-EPI creatinine equation, warns that the result is less reliable at extremes of muscle mass, in bodybuilders or in people with muscle-wasting conditions, and advises adults not to eat any meat in the 12 hours before an eGFR test.

ALT is the line where printed ceilings vary most. British guidance defines an abnormal liver blood test simply as a value outside the laboratory’s reference interval, while noting growing evidence that current ALT limits may be set too high. The Leeds laboratory prints 40 U/L as its adult ceiling; others print different figures, so the same ALT can be flagged in one place and not in another.

Ferritin carries a trap. It reflects iron stores but also rises with inflammation, so a reassuring ferritin during an infection or an inflammatory illness can hide low iron. British guidance treats a ferritin below 15 µg/L as absent iron stores and below 30 µg/L as generally low, and suggests 45 µg/L as the most useful practical cut-off for iron deficiency.

TestTypical UK rangeNotes
Fasting glucose3.5 – 6.0 mmol/LFasting only. 6.1–6.9 is impaired fasting glucose.
HbA1cbelow 42 mmol/molAverage over about three months. No fasting needed.
Creatinine62 – 115 µmol/L (m) · 49 – 90 (f)Depends on muscle mass. Read with eGFR.
eGFR90 or more mL/min/1.73 m²A calculation. 60–89 alone, with no other kidney signs, is not CKD.
ALTup to 40 U/LCeiling varies by laboratory. Read with ALP, bilirubin and GGT.
TSH0.2 – 4.3 mU/LVaries by laboratory and through the day. Prefer a morning test.
Ferritin30 – 337 µg/LAnother NHS lab: 24–340 (m), 11–310 (f). Inflammation can mask low iron.
Cholesterolno single rangeNICE uses your overall risk. A fasting sample is not needed.

Ranges, thresholds and targets

Not every number in a chart like this is a reference range, and mixing them up causes a lot of needless worry. Four kinds of figure appear on UK results, and each answers a different question.

HbA1c needs no fasting, and NICE’s lipid guideline states that a fasting sample is not needed for a full lipid profile either. In England, the NHS Health Check offered every five years to people aged 40 to 74 includes a cholesterol test and sometimes a blood sugar test, which is where many adults first meet these thresholds.

42–47
HbA1c in mmol/mol that NICE treats as high risk of type 2 diabetes.[9]
≥ 48
HbA1c in mmol/mol at which NICE suspects type 2 diabetes, to be confirmed by a repeat test.[9]
> 40 %
Fall in non-HDL cholesterol NICE aims for when a statin is used for prevention.[10]
  1. Reference rangeWhere most healthy people fall on that laboratory’s method. Sitting just outside it is common in healthy people.
  2. Diagnostic thresholdA cut-off agreed in guidance. NICE treats an HbA1c of 48 mmol/mol or more, or a fasting glucose of 7.0 mmol/L or more, as possible type 2 diabetes, to be confirmed by a second test.
  3. Treatment targetA goal tied to your risk. For people with heart or circulatory disease, NICE monitors an LDL of 2.0 mmol/L or less, or a non-HDL of 2.6 mmol/L or less.
  4. Critical valueA result so far out that UK pathology guidance asks the laboratory to phone it to the clinician, usually within 2 hours.

What shifts a range

Sex and age change several lines outright. Haemoglobin limits are similar in boys and girls until adolescence and then separate. Pregnancy lowers them again: UK guidance defines anaemia in pregnancy as a haemoglobin below 110 g/L in the first trimester and below 105 g/L after that. Children have their own values for many lines, so a child’s results should never be read against an adult chart like this one.

Two more shifts are invisible on your results. The first is method: changing laboratories in the middle of a follow-up can move a result without anything changing in you. The second is the clock. TSH is highest in the early morning, and in one large hospital data set it was up to about a quarter lower by early afternoon than at 7 am, which is why a morning test makes results easier to compare.

Preparation is the third. European recommendations on blood sampling treat patient preparation, fasting included, as part of a valid result. If a flagged value could have been moved by a meal, the time of day or a change of laboratory, it is reasonable to ask whether it should be repeated under the same conditions.

Optimal ranges and red flags

Charts that promise optimal ranges answer a different question. A reference range describes healthy people, while some thresholds are chosen for a purpose and can differ for the same test. The family medicine review gives folate as an example: one level prevents anaemia, a higher one is needed for homocysteine to settle. Neither is your laboratory’s range, and neither should drive treatment without your GP.

Real red flags are less about one letter than about patterns: a value far outside its range, several related lines moving the same way, a clear change from your own previous result, or an abnormal result alongside symptoms. For results that are dangerously far out, Royal College of Pathologists guidance asks laboratories to phone the clinician who requested the test, usually within 2 hours; a potassium of 6.5 mmol/L or more and a sodium of 120 mmol/L or less are among its examples. The most urgent findings are designed to reach a professional without depending on you.

A longer panel is not a safer one. A Cochrane review of general health checks in 251,891 adults found little or no effect on deaths, while every extra line adds one more chance of a flag to explain.

What a good explanation looks like

A value, a range and a flag leave you with the hardest part: deciding whether it matters. The family medicine review argues that results should say so in plain words: no immediate concern, discuss at your next visit, or contact your doctor soon. In England, the short comment a GP adds to a result, such as ‘normal’ or ‘satisfactory’, plays a similar role, though the wording differs from one surgery to another.

The format helps too. A 2024 systematic review of 18 studies found that showing a result as a horizontal bar against its range improved understanding, and made people less inclined to search online or contact their doctor than numbers with a range alone. Most sets of results still offer neither, so the useful questions fall to you.

Take four of them to your appointment: how far outside is the value, in which direction, which related lines agree with it, and how does it compare with your last result? They turn a general chart like this one into a conversation about you.