Blood Sugar Levels Chart: Normal Glucose Levels & Ranges
Blood sugar levels chart in mg/dL — fasting, after eating, random and A1C, with normal, prediabetes and diabetes ranges. What is a normal glucose level?
Blood glucose — blood sugar — is the amount of sugar circulating in your blood, and the blood glucose test is one of the most frequently ordered labs in the United States. This guide gives you the blood sugar levels chart you came for, in mg/dL, and explains what each number means: what a normal glucose level is, where prediabetes starts (and why that line is disputed), what counts as a dangerous blood sugar, why your reading climbs at dawn when you haven't eaten, and what works to bring a high level down.
Key takeaways
- A normal fasting blood sugar is 99 mg/dL or below; labs commonly print 70–99 mg/dL.12
- Prediabetes = fasting 100–125 mg/dL, A1C 5.7%–6.4%, or a 2-hour OGTT of 140–199 mg/dL.1
- Diabetes = fasting ≥ 126 mg/dL, A1C ≥ 6.5%, or a 2-hour/random glucose ≥ 200 mg/dL — usually confirmed by a second test.13
- The prediabetes line is not universal. The ADA starts it at 100 mg/dL; the WHO and IDF at 110 mg/dL. Same blood, different label.45
- The thresholds don't move with age. 126 mg/dL fasting means diabetes at 30 and at 80.3
- Prediabetes is common and reversible: 97.6 million U.S. adults had it in 2021, and lifestyle change cut progression to diabetes by 58%.67
- A low under 70 mg/dL, or a very high glucose with symptoms, needs attention now.89
Blood sugar levels chart
Here is the chart, for adults who are not pregnant. All values are plasma glucose in mg/dL, the unit on U.S. lab reports, as published by the NIDDK (NIH).1
| Test | Normal | Prediabetes | Diabetes |
|---|---|---|---|
| Fasting plasma glucose (FPG) | 99 mg/dL or below | 100 – 125 mg/dL | 126 mg/dL or above* |
| A1C | below 5.7% | 5.7% – 6.4% | 6.5% or above* |
| 2-hour OGTT (75 g) | 139 mg/dL or below | 140 – 199 mg/dL | 200 mg/dL or above* |
| Random plasma glucose | — | — | 200 mg/dL or above (with symptoms) |
* A diabetes diagnosis is normally confirmed with a second test — a repeat on a new sample, or two different tests. The exception is unmistakable hyperglycemia symptoms with a clearly elevated glucose.13
On the "normal" row: NIDDK states the normal fasting result as "99 mg/dL or below" and gives no floor. Most U.S. labs print 70–99 mg/dL (3.9–5.5 mmol/L), using 70 because that's the conventional alert value for hypoglycemia.2
Converting units. The U.S. reports glucose in mg/dL; most of the world uses mmol/L, and France also uses g/L. The math: mg/dL ÷ 18 = mmol/L, and g/L × 100 = mg/dL. So 126 mg/dL is 7.0 mmol/L — and 1.26 g/L on a French report. Same result, three units.
The key thresholds in all three:
| mg/dL (U.S.) | mmol/L | g/L | |
|---|---|---|---|
| Bottom of normal / low sugar alert | 70 | 3.9 | 0.70 |
| Top of normal, fasting | 99 | 5.5 | 0.99 |
| Prediabetes starts (ADA) | 100 | 5.6 | 1.00 |
| Prediabetes starts (WHO) | 110 | 6.1 | 1.10 |
| Diabetes, fasting | 126 | 7.0 | 1.26 |
| Normal ceiling, 2 h after eating | 140 | 7.8 | 1.40 |
| Diabetes, 2-hour or random | 200 | 11.1 | 2.00 |
| Clinically significant low | 54 | 3.0 | 0.54 |
Why the prediabetes line is disputed
Most articles print 100 mg/dL and move on. The number is a choice, not a fact of nature.
The American Diabetes Association defines impaired fasting glucose as 100–125 mg/dL. The World Health Organization and the International Diabetes Federation define it as 110–125 mg/dL.45 The ADA originally used 110 too, and lowered the cutoff to 100 in 2003 to catch more people at risk earlier.5
The consequence isn't academic. If your fasting glucose is 104 mg/dL, you have prediabetes in the United States and you don't in much of the rest of the world. Nothing about your body changed — only the ruler.5
Which line is right? There's a real argument for the lower one: in one European study, the ADA's 100 mg/dL definition identified additional people whose cardiovascular risk was genuinely elevated — people the WHO cutoff called normal.4 Catching them earlier is the point, since prediabetes is where the trajectory can still be bent.
Prediabetes also has a weakness the guidelines acknowledge: it is less reproducible than diabetes, reproducing on retest about 50% of the time versus roughly 70%.5 A single borderline value is a soft signal — an argument for rechecking it, not ignoring it.
What this means for you: between 100 and 110 mg/dL, the label depends on who's reading. Recheck it, look at your A1C alongside it, and treat it as an early, reversible signal.
What is blood glucose?
Glucose is your body's main fuel. It comes mostly from the carbohydrates you eat and powers your cells, your muscles, and above all your brain.
To keep the supply steady, your pancreas balances two hormones. Insulin lowers blood sugar by moving glucose into cells; glucagon raises it by pulling stored glucose out of the liver. When that balance breaks — too little insulin, or cells no longer responding to it (insulin resistance) — glucose stays in the blood and levels climb. That's the machinery behind type 2 diabetes.10
Why is blood glucose measured?
Your primary care provider may order a glucose test to:1112
- screen for diabetes and prediabetes — the USPSTF recommends screening all adults aged 35 to 70 with overweight or obesity (BMI ≥ 25, or ≥ 23 for Asian American adults), every 3 years if normal. It's a Grade B recommendation, so insurers generally cover it without cost sharing;12
- investigate symptoms: excessive thirst, frequent urination, fatigue, weight loss, blurred vision;
- monitor known diabetes and its treatment;
- as part of a routine basic (BMP) or comprehensive metabolic panel (CMP), alongside creatinine and BUN;
- during pregnancy, to screen for gestational diabetes — usually a glucose challenge test between weeks 24 and 28.11
That screening matters more than it sounds: 8.7 million U.S. adults have diabetes and don't know it — 22.8% of everyone who has it.6 This is a disease found by testing, not by feeling sick.
How the test is done
A fasting plasma glucose is drawn from a vein after at least 8 hours of fasting. Water is fine — staying hydrated actually makes the draw easier. Coffee (yes, even black), tea, and smoking should be avoided during the fast, since they can shift the result. The day before, eat normally: a fasting glucose is only useful if it reflects your ordinary metabolism.
Several complementary ways exist to look at glucose:113
- Fasting plasma glucose (FPG) — the workhorse screening test, 8 hours fasting;
- Random plasma glucose — any time, no fasting; useful when symptoms are obvious;
- A1C — no fasting needed, reflects average glucose over roughly 3 months;
- Oral glucose tolerance test (OGTT) — measured 2 hours after drinking 75 g of glucose; more sensitive than FPG;
- Glucose challenge test — the shorter, non-fasting version used in pregnancy;
- Continuous glucose monitoring (CGM) — a skin sensor, central to modern diabetes care.13
Normal blood glucose levels
So the short answer to what is a normal glucose level: 70 to 99 mg/dL fasting, and under 140 mg/dL two hours after eating.12 After a meal glucose rises — that's supposed to happen — and in someone without diabetes it's generally back below 140 within two hours, which is why 140 is the OGTT normal ceiling.1 A level that stays high long after eating is worth checking. For people with diabetes, post-meal targets are set individually; the range used for continuous monitoring, 70 to 180 mg/dL, reflects roughly where most adults with diabetes aim.9
Good to know: reference ranges vary slightly between laboratories depending on the assay. The range that applies to you is the one printed on your report.
Does normal blood sugar change with age?
A much-searched question, usually phrased as "normal blood sugar for a 60-year-old." The answer that matters: the diagnostic thresholds are identical at every age.3 There is no higher normal that becomes acceptable because you got older. 126 mg/dL fasting means diabetes at 25 and at 85.
Two nuances. Glucose does drift slightly upward with age across the population — but drifting toward a threshold isn't the threshold moving. And in an older, frail adult who already has diabetes, treatment targets are deliberately individualized and often loosened, because a hypoglycemic episode is more dangerous than a slightly high A1C. What flexes is the management goal, never the diagnostic line.
Interpreting your results
High blood sugar (hyperglycemia)
A glucose above the normal range can mean prediabetes or diabetes — but not always. Acute stress, infection, surgery, steroids like prednisone, and simply a fast that wasn't respected can all push it up temporarily. This is why a single result never establishes a diagnosis: it gets rechecked and read in context.3
Low blood sugar (hypoglycemia)
Below 70 mg/dL is the threshold NIDDK uses for low blood glucose.8 Symptoms include shakiness, sweating, sudden hunger, a racing heartbeat, headache, dizziness, and trouble concentrating. In someone without diabetes it's uncommon: prolonged fasting, hard exercise, alcohol, certain medications. In someone treated for diabetes it's usually the treatment — insulin or a sulfonylurea — and needs a conversation with the prescriber, not a solo dose adjustment.
Dangerous blood sugar levels: when to worry
There's no single "dangerous number," but two situations call for acting quickly:
- Low. Under 54 mg/dL (3.0 mmol/L) hypoglycemia is clinically significant — low enough to matter regardless of symptoms.9 Severe hypoglycemia — confusion, seizure, loss of consciousness, or any low you can't treat yourself — is an emergency.8
- High. A glucose above roughly 250 mg/dL (level 2 hyperglycemia)9 with symptoms — intense thirst, heavy urination, nausea, abdominal pain, drowsiness, rapid breathing, or fruity-smelling breath — can signal diabetic ketoacidosis or a hyperosmolar state. That's an emergency room, not a next-week appointment.
Outside those extremes, an abnormal number is not an emergency. It's a result to interpret calmly, with the rest of your panel.
Why does my blood sugar go up in the morning without eating?
Because your liver works the night shift. In the early morning, your body releases hormones — growth hormone, cortisol — that push the liver to produce glucose in preparation for waking up. This is the dawn phenomenon, and it explains a fasting reading higher than expected with no food to blame. It's not rare: in both type 1 and type 2 diabetes, prevalence exceeds 50%, and it can raise A1C by as much as 0.4%.14
A related but distinct pattern is the Somogyi effect, or post-hypoglycemic rebound — a low overnight triggering hormones that overshoot into a morning high. The distinction: the dawn phenomenon is not preceded by hypoglycemia, and is considerably more common.14 The two call for opposite responses, which is why morning highs get investigated rather than assumed.
Other things raise glucose "for no reason": poor sleep, stress, a brewing infection, some medications. If it repeats, bring it to your clinician — patterns are read over time, not off one reading.
How to lower blood sugar
Several lifestyle levers genuinely move a high glucose. They complement medical care rather than replace it: if you're on treatment, never change or stop it on your own.
- Move — especially after meals. A meta-analysis found exercise blunts the post-meal spike more effectively after the meal than before it.15 Even 10 to 15 minutes of walking helps.
- Change what's on the plate. Favor high-fiber foods (vegetables, legumes, whole grains); cut sugar-sweetened beverages and refined sugars.
- Lose weight if you carry extra. Even modest loss improves glucose, and in prediabetes can bring you back to normal — proportionally to how much you lose.16
- Sleep enough and manage stress. Both raise glucose through the same hormones as the dawn phenomenon.
What about "natural remedies" — cinnamon, herbal teas? An umbrella review pooling eleven meta-analyses found cinnamon does reduce fasting glucose and A1C — but by around 0.1% of A1C.17 That's a rounding error next to walking after dinner. A possible add-on, not a treatment.
"How do I lower my blood sugar fast before a blood test?" You don't. The point is to measure your usual glucose. Gaming the result makes the number useless and can delay a diagnosis that would have helped you.
Prediabetes: a warning you can act on
Prediabetes is not a sentence — it's a window. 97.6 million U.S. adults, more than 1 in 3, had it in 2021, and most don't know it.6
Now the good news. The Diabetes Prevention Program, the landmark U.S. trial, showed a lifestyle program cut progression to type 2 diabetes by 58% over 3 years — and by 71% in participants aged 60 and older, the group usually assumed to be past helping. Metformin cut it by 31%. The dose-response was clean: every kilogram lost brought a 16% reduction in risk.7 NIDDK frames the goal as losing 5% to 7% of your starting weight.18
And it lasted: incidence was still reduced 34% at 10 years and 27% at 15 years.7 A meta-analysis of 44 trials confirms the shape — weight loss of even 1% to 9% raises the probability of returning to normal glucose, in proportion to how much is lost.16 Every pound counts, and it is never too early to start.
Diabetes: thresholds, confirmation, and remission
Diabetes is diagnosed on a fasting glucose ≥ 126 mg/dL, an A1C ≥ 6.5%, or a glucose ≥ 200 mg/dL — 2 hours into an OGTT or at random with classic symptoms — confirmed with a second test.1193
Type 2 diabetes shares its roots — excess weight, insulin resistance — with metabolic dysfunction-associated steatotic liver disease (MASLD), the fatty liver condition. They travel together, which is why diabetes often comes with a raised ALT.
The headline of the last decade: type 2 diabetes can go into remission. In the UK DiRECT trial, an intensive primary-care weight-management program achieved remission — A1C below 6.5% with no diabetes medication — in 46% of participants at 1 year,20 with more than a third sustained at 2 years, durability tracking with keeping the weight off.21 It works best in recent-onset diabetes and must be medically supervised — but it changed how the disease is understood: as something that can move in both directions.
Beyond A1C: time in range
A1C remains the monitoring reference, but it only gives an average — and an average conceals its own volatility. Two people with an identical 7.0% can be living completely different days.
With CGM sensors now widespread, a second metric has taken hold: time in range (TIR), the share of the day spent between 70 and 180 mg/dL. International consensus recommends most adults with diabetes spend more than 70% of the day in that window, while keeping time below 70 mg/dL under 4% and below 54 mg/dL under 1%.9 That last number is the point: TIR doesn't just reward lower glucose, it penalizes the lows an A1C hides completely.
Sensors help: two meta-analyses of randomized trials in type 2 diabetes show a modest but real A1C reduction — around 0.2 to 0.3% — versus fingerstick monitoring.1322 The field is refining further with time in tight range (TITR), 70–140 mg/dL, across over 20,000 CGM users.23
What can affect your blood glucose
Plenty of things move the number with no disease involved: food and whether the fast was observed; recent exercise; stress or acute illness; medications, especially corticosteroids; thyroid imbalance, since thyroid hormones help regulate glucose metabolism — an unexplained glucose can justify checking a TSH; pregnancy, which changes the thresholds; and anemia or a hemoglobin variant, which can distort A1C — worth knowing if your hemoglobin or MCV is off.
That last one deserves emphasis: A1C has real limits. Because it measures glucose stuck to hemoglobin, anything changing red cell lifespan changes A1C independently of your glucose — with anemia or certain hemoglobin variants, it can underestimate true glucose. The proposed alternatives, glycated albumin and fructosamine, do not improve accuracy in exactly the patients whose A1C is unreliable.24 Hence the rule governing this page: cross-reference several markers rather than lean on one number.
When to see a doctor
Talk to your provider if your glucose falls outside the normal range — especially with symptoms: unusual thirst, frequent urination, marked fatigue, unexplained weight loss, blurred vision. If your fasting glucose sits in the 100–125 mg/dL prediabetes band, that's a conversation to have now, while the odds are best. Seek care immediately for severe hypoglycemia or a very high glucose with symptoms.
Get your blood glucose interpreted by AI DiagMe
A glucose is never read alone: what it means depends on your other markers and your profile — age, weight, medications, A1C. The same 108 mg/dL means something different in two different people.
👉 AI DiagMe interprets your lab results — blood, urine, or stool — in plain language, taking your whole context into account. An informational service that does not provide a diagnosis and complements, never replaces, your physician.
Recent research
According to recent publications indexed on PubMed:
- Monitoring is getting personal. CGM sensors and time in range now complement A1C, with consensus targets (>70% between 70–180 mg/dL) and real-world data on over 20,000 users pushing toward a tighter 70–140 mg/dL window.923
- Lifestyle carries more weight than expected. Weight loss linearly predicts return to normal glucose in prediabetes,16 remission of recent-onset type 2 diabetes is a realistic primary-care goal,2021 and walking after meals beats before.15
- "Remedies" are better measured, and the tests' limits are clearer. For cinnamon, an umbrella review of eleven meta-analyses confirms a real but small effect — an adjunct, not a therapy.17 And we better understand when A1C misleads (anemia, hemoglobin variants) — and that glycated albumin and fructosamine don't rescue it.24
These findings concern monitoring and management; they do not change the diagnostic thresholds.
Frequently asked questions
What is a normal glucose level?
At what blood sugar level are you diabetic?
What blood sugar level is dangerous?
What are prediabetes numbers?
How do I lower my blood sugar?
Does cinnamon lower blood sugar?
Why is my blood sugar high in the morning without eating?
Do I need to fast? Can I drink coffee?
Does normal blood sugar change with age?
How do I convert mg/dL to mmol/L?
What's the difference between blood glucose and A1C?
Is prediabetes reversible? Can type 2 diabetes be cured?
Bottom line
Remember the chart: 70–99 mg/dL normal fasting, 100–125 prediabetes, 126 or above diabetes, with 5.7% and 6.5% as the matching A1C lines — and these thresholds hold at every age. Remember that the prediabetes line is a judgment call (100 in the U.S., 110 per the WHO), that prediabetes is common and reversible, and that no single number is ever a diagnosis. For a reliable reading, your full set of markers and your profile have to be taken into account — exactly what AI DiagMe does, alongside your physician.
Sources
Official U.S. sources and peer-reviewed publications (PubMed) used for this guide:
Footnotes
-
National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK, NIH) — Diabetes Tests & Diagnosis. niddk.nih.gov ↩ ↩2 ↩3 ↩4 ↩5 ↩6 ↩7 ↩8
-
Cleveland Clinic — Blood Glucose Test. my.clevelandclinic.org ↩ ↩2 ↩3
-
Sacks DB, et al. Guidelines and Recommendations for Laboratory Analysis in the Diagnosis and Management of Diabetes Mellitus. Diabetes Care, 2023. PubMed · DOI ↩ ↩2 ↩3 ↩4 ↩5 ↩6 ↩7
-
Filippatos TD, Rizos EC, Gazi IF, et al. Differences in metabolic parameters and cardiovascular risk between American Diabetes Association and World Health Organization definition of impaired fasting glucose in European Caucasian subjects: a cross-sectional study. Arch Med Sci, 2013;9(5):788-795. PubMed · DOI ↩ ↩2 ↩3
-
Bansal N. Prediabetes diagnosis and treatment: A review. World J Diabetes, 2015;6(2):296-303. PubMed · DOI ↩ ↩2 ↩3 ↩4 ↩5
-
NIDDK (NIH) — Diabetes Statistics. niddk.nih.gov ↩ ↩2 ↩3
-
NIDDK (NIH) — Evidence Supporting Prevention (Diabetes Prevention Program and DPP Outcomes Study results). niddk.nih.gov ↩ ↩2 ↩3
-
NIDDK (NIH) — Low Blood Glucose (Hypoglycemia). niddk.nih.gov ↩ ↩2 ↩3
-
Battelino T, et al. Clinical Targets for Continuous Glucose Monitoring Data Interpretation: Recommendations From the International Consensus on Time in Range. Diabetes Care, 2019. PubMed · DOI ↩ ↩2 ↩3 ↩4 ↩5 ↩6
-
Khan RMM, et al. From Pre-Diabetes to Diabetes: Diagnosis, Treatments and Translational Research. Medicina (Kaunas), 2019. PubMed · DOI ↩
-
MedlinePlus (U.S. National Library of Medicine, NIH) — Blood Glucose Test. medlineplus.gov ↩ ↩2 ↩3
-
U.S. Preventive Services Task Force — Prediabetes and Type 2 Diabetes: Screening, 2021 (Grade B). uspreventiveservicestaskforce.org ↩ ↩2
-
Jancev M, et al. Continuous glucose monitoring in adults with type 2 diabetes: a systematic review and meta-analysis. Diabetologia, 2024. PubMed · DOI ↩ ↩2
-
Rana MA, et al. Dawn Phenomenon. StatPearls, NCBI Bookshelf, 2023. Bookshelf ID NBK430893. bookshelf ↩ ↩2
-
Engeroff T, et al. After Dinner Rest a While, After Supper Walk a Mile? A Systematic Review with Meta-analysis on the Acute Postprandial Glycemic Response to Exercise. Sports Med, 2023. PubMed · DOI ↩ ↩2
-
Jayedi A, et al. Efficacy of lifestyle weight loss interventions on regression to normoglycemia and progression to type 2 diabetes in individuals with prediabetes: a systematic review and meta-analyses. Am J Clin Nutr, 2024. PubMed · DOI ↩ ↩2 ↩3
-
Zarezadeh M, et al. The effect of cinnamon supplementation on glycemic control: an umbrella meta-analysis on interventional meta-analyses. Diabetol Metab Syndr, 2023. PubMed · DOI ↩ ↩2
-
NIDDK (NIH) — Insulin Resistance & Prediabetes. niddk.nih.gov ↩
-
Sapra A, Bhandari P. Diabetes. StatPearls, NCBI Bookshelf, 2023. Bookshelf ID NBK551501. bookshelf ↩
-
Lean MEJ, et al. Primary care-led weight management for remission of type 2 diabetes (DiRECT): an open-label, cluster-randomised trial. Lancet, 2018. PubMed · DOI ↩ ↩2
-
Lean MEJ, et al. Durability of a primary care-led weight-management intervention for remission of type 2 diabetes: 2-year results of the DiRECT trial. Lancet Diabetes Endocrinol, 2019. PubMed · DOI ↩ ↩2
-
Seidu S, et al. Efficacy and Safety of Continuous Glucose Monitoring and Intermittently Scanned Continuous Glucose Monitoring in Patients With Type 2 Diabetes. Diabetes Care, 2024. PubMed · DOI ↩
-
Dunn TC, et al. Is It Time to Move Beyond TIR to TITR? Real-World Data from Over 20,000 Users of Continuous Glucose Monitoring. Diabetes Technol Ther, 2024. PubMed · DOI ↩ ↩2
-
Niwaha AJ, et al. Glycated albumin and fructosamine do not improve accuracy of glycaemic control assessment in patients with conditions reported to affect HbA1c reliability. Diabet Med, 2025. PubMed · DOI ↩ ↩2