There is no single “normal blood glucose level” that applies to every measurement because glucose naturally changes according to whether a person is fasting, has recently eaten, is pregnant, has diabetes, is taking glucose-lowering treatment or is experiencing illness or exercise. For non-pregnant adults without diabetes, fasting plasma glucose below 5.6 mmol/L is within the normal range under ADA criteria, while 5.6–6.9 mmol/L is classified as impaired fasting glucose or prediabetes and 7.0 mmol/L or above meets the diabetes threshold when confirmed appropriately.
A two-hour glucose below 7.8 mmol/L during a 75-g oral glucose tolerance test is generally considered normal, 7.8–11.0 mmol/L indicates impaired glucose tolerance and 11.1 mmol/L or above meets the diabetes criterion. These are diagnostic categories, not necessarily the same as targets used by people already diagnosed with diabetes.
In everyday UK self-monitoring guidance, healthy blood glucose targets are often expressed differently from laboratory diagnostic thresholds. Diabetes UK gives a general target of 4–7 mmol/L on waking before eating and 4–7 mmol/L before meals for adults living with diabetes, while emphasising that individual targets can differ. NHS-derived guidance similarly lists 4–7 mmol/L before meals as a general target for many adults with type 2 diabetes and gives post-meal targets that vary by diabetes type.
These are treatment targets rather than definitions of “normal physiology” in someone without diabetes. A person without diabetes can experience temporary post-meal glucose values above a fasting range without having diabetes, while a person with diabetes may have an individually agreed target that differs from a generic range. The words “normal”, “healthy”, “target” and “diagnostic threshold” therefore should not be treated as synonyms.
Fasting glucose is one of the clearest contexts for defining a normal laboratory value. Fasting means that no caloric intake has occurred for at least eight hours, although water is normally permitted. Under ADA diagnostic criteria, fasting plasma glucose below 5.6 mmol/L is below the prediabetes threshold, 5.6–6.9 mmol/L represents impaired fasting glucose, and 7.0 mmol/L or higher meets the diabetes threshold when confirmed if necessary.
Other organisations and countries have historically used slightly different thresholds for impaired fasting glucose, which is why published “normal ranges” can appear inconsistent. WHO materials, for example, have used 6.1 mmol/L as the lower boundary for impaired fasting glycaemia in some diagnostic frameworks. The difference illustrates why the laboratory's reference range and the guideline being used should be identified rather than assuming one universal cutoff exists worldwide.
After eating, blood glucose normally rises because digested carbohydrate enters the bloodstream. Insulin secretion increases, tissues take up glucose and the liver changes from net glucose release to storage, so glucose should subsequently fall toward the person's baseline. The oral glucose tolerance test provides a standardised way to assess this response: after a 75-g glucose load, a two-hour plasma glucose below 7.8 mmol/L is generally classified as normal, 7.8–11.0 mmol/L as impaired glucose tolerance and 11.1 mmol/L or higher as diabetes.
Ordinary meals are not equivalent to the standardised glucose solution used in an OGTT, so a random home glucose reading two hours after an ordinary meal should not automatically be compared directly with the OGTT diagnostic threshold. The amount and type of carbohydrate consumed, meal composition, physical activity and individual insulin sensitivity all affect post-meal readings.
HbA1c provides another way to describe normal and abnormal glucose regulation without requiring a particular time of day. An HbA1c below 5.7% is generally below the ADA prediabetes range, 5.7–6.4% indicates prediabetes and 6.5% or above meets the diagnostic criterion for diabetes when the test is appropriate and confirmed where required. HbA1c is reported in the UK in both percentage and mmol/mol, with 5.7% corresponding to approximately 39 mmol/mol and 6.5% to 48 mmol/mol.
HbA1c is an average-related measure rather than a direct glucose concentration, so it cannot tell whether glucose has remained stable or has fluctuated substantially. Conditions affecting red-cell lifespan, haemoglobin structure or other factors can also make HbA1c less reliable, in which circumstances plasma glucose criteria may be preferred.
A normal glucose level is different from an appropriate diabetes treatment target. For example, Diabetes UK states that a general healthy target for adults living with diabetes is 4–7 mmol/L when waking and before meals, but explicitly notes that individual targets can differ. NHS-derived guidance lists different targets according to diabetes type, age and pregnancy status.
For adults with type 1 diabetes, examples include 4–7 mmol/L before meals and 5–9 mmol/L two hours after eating, whereas type 2 guidance may use a two-hour post-meal target below 8.5 mmol/L. During pregnancy, targets are stricter, with ADA guidance recommending fasting values of 3.9–5.3 mmol/L and one-hour post-meal values below 7.8 mmol/L in relevant diabetes-in-pregnancy groups. Treatment targets are chosen to balance benefits of reducing hyperglycaemia against the immediate dangers of hypoglycaemia.
Blood glucose below the usual healthy range is also clinically important. In diabetes care, a glucose below 3.9 mmol/L is commonly treated as hypoglycaemia requiring attention, particularly in people using insulin or medicines capable of causing low glucose. Symptoms can include shaking, sweating, hunger, palpitations, weakness, difficulty concentrating and confusion. More severe hypoglycaemia can cause seizures, unconsciousness and potentially death if untreated.
Conversely, a single high glucose result does not automatically establish diabetes because glucose can temporarily rise during meals, stress, acute illness or treatment with certain medicines. Persistent abnormal results require appropriate diagnostic assessment. The interpretation of any individual value therefore depends on the measurement method, timing relative to food, symptoms, medications, pregnancy status and whether the person has an established diagnosis.
The most useful answer to “what is a normal blood glucose level?” is therefore conditional: for a non-pregnant person without diabetes, fasting plasma glucose below 5.6 mmol/L and a two-hour 75-g OGTT value below 7.8 mmol/L are commonly used as normal categories under ADA criteria; for everyday diabetes management, many UK adults are given pre-meal targets around 4–7 mmol/L, while pregnancy uses narrower targets. These numbers should not be mixed together because they answer different questions.
Diagnostic thresholds establish whether diabetes is present; treatment targets guide management after diagnosis; reference intervals describe laboratory populations; and personal targets are adjusted according to clinical circumstances. If a home glucose result is repeatedly outside the expected range, particularly if accompanied by thirst, frequent urination, unexplained weight loss, vomiting or other concerning symptoms, formal medical assessment is appropriate rather than relying on a single home reading.
