Glucose is measured through blood tests, either from a finger prick, a lab blood draw, or a small sensor worn under the skin. The most common methods are fasting blood glucose tests, A1C tests that reflect a two- to three-month average, and continuous glucose monitors that track levels around the clock. Each approach measures something slightly different, and the right one depends on whether you need a single snapshot or an ongoing picture of how your body handles sugar.
Finger-Prick Glucose Meters
The most familiar way to measure glucose is with a portable meter and a disposable test strip. You lance the side of your fingertip, touch a small drop of blood to the strip, and get a reading in about five seconds. Inside the strip, an enzyme reacts with the glucose in your blood and strips away electrons from each glucose molecule. A second chemical, called a mediator, ferries those electrons to a tiny electrode. The meter measures the resulting electrical current and converts it into a number in mg/dL.
Older strips used an enzyme called glucose oxidase, but most modern strips have switched to dehydrogenase-based enzymes. The practical difference: dehydrogenase enzymes aren’t thrown off by oxygen in the blood sample, which makes them more reliable in everyday use.
Fingertips are the preferred testing site because blood flow there is high, so glucose levels closely match what’s circulating in your body at that moment. You can test on the palm, forearm, or thigh, but readings from these sites can lag behind fingertip readings by about 20 minutes. That lag doesn’t matter much when glucose is stable, but if your levels are rising quickly after a meal or dropping during exercise, an alternative site may give you a misleadingly calm number.
Lab Tests Your Doctor Orders
Three standard lab tests are used to screen for and diagnose diabetes. Each has specific cutoff numbers that separate normal results from prediabetes and diabetes.
Fasting Plasma Glucose
You fast for at least eight hours, then have blood drawn. A result of 99 mg/dL or below is normal. Between 100 and 125 mg/dL indicates prediabetes. At 126 mg/dL or above, the result points to diabetes. This test captures your baseline glucose level when your body hasn’t had to process any food.
Oral Glucose Tolerance Test
This test checks how efficiently your body clears sugar from the bloodstream. After fasting, you drink a sugary solution, then have your blood drawn two hours later. A reading of 139 mg/dL or below is normal. Between 140 and 199 mg/dL falls in the prediabetes range. A result of 200 mg/dL or above indicates diabetes. It’s more time-consuming than a simple fasting test but better at catching early insulin resistance that a fasting number might miss.
A1C (Glycated Hemoglobin)
The A1C test doesn’t measure the glucose in your blood right now. Instead, it measures how much sugar has attached to hemoglobin, the protein inside red blood cells that carries oxygen. Because red blood cells live roughly 90 to 120 days, the A1C reflects your average blood sugar over the past two to three months. No fasting required.
An A1C below 5.7% is normal. Between 5.7% and 6.4% signals prediabetes. At 6.5% or above, it meets the threshold for diabetes. The test is convenient and gives a broader view than any single glucose reading, but it has blind spots. Conditions that shorten the lifespan of red blood cells, such as certain anemias or chronic kidney disease, can pull the A1C number artificially low, making blood sugar control look better than it actually is. Differences in individual hemoglobin glycation rates can also skew results. If your doctor suspects these factors are at play, they’ll rely more on direct glucose measurements.
Continuous Glucose Monitors
A continuous glucose monitor, or CGM, is a small sensor inserted just beneath the skin, usually on the back of the upper arm or the abdomen. It measures glucose in the interstitial fluid (the thin layer of fluid surrounding your cells) every one to five minutes and sends the data wirelessly to a phone or receiver. Instead of isolated snapshots, you get a continuous curve showing how your glucose rises after meals, dips during exercise, and behaves overnight.
Because the sensor reads interstitial fluid rather than blood directly, there is a time lag. Readings can trail actual blood glucose by anywhere from a few minutes to as much as 50 minutes, though under typical conditions the delay is shorter. This matters most when glucose is changing rapidly. If you feel symptoms of a low and your CGM still shows a normal number, a finger prick is the faster confirmation.
Accuracy is measured by a statistic called MARD, or mean absolute relative difference. Lower is better. The Dexcom G7 reports a MARD of 8.2% in adults, and the FreeStyle Libre 3 comes in at 7.9% to 8.9% depending on study conditions. Both are accurate enough to guide insulin dosing decisions for most people, though no CGM perfectly matches a lab blood draw at every moment.
Time in Range: A Newer Way to Use CGM Data
CGMs generate so much data that a single average doesn’t capture the full picture. That’s why clinicians and patients increasingly focus on “time in range,” which is the percentage of the day your glucose stays between 70 and 180 mg/dL. The American Diabetes Association recommends aiming for at least 70% of readings in that window, which works out to roughly 17 out of 24 hours. Some people, like those who are pregnant or elderly, may have tighter or looser targets.
Time in range is useful because two people can have the same A1C but very different daily patterns. One might hold steady at 140 mg/dL most of the day. Another might swing between 60 and 250 mg/dL and land on the same average. Time in range reveals those swings in a way that A1C cannot.
How Often You Need to Test
Testing frequency depends on your situation. People with type 1 diabetes or type 2 diabetes who take insulin typically check multiple times a day, often before meals, after meals, before bed, and before physical activity. If you have type 2 diabetes managed with oral medications or lifestyle changes alone, your doctor may recommend less frequent checks, sometimes just a few times a week or primarily in the morning.
A CGM reduces the burden of manual testing since it tracks glucose automatically. But even CGM users occasionally need a finger prick to confirm a reading when the sensor result doesn’t match how they feel, or when glucose is changing fast.
Non-Invasive Methods in Development
Several technologies aim to measure glucose without breaking the skin at all. Near-infrared spectroscopy shines light through the skin and analyzes how glucose absorbs specific wavelengths, but glucose’s absorption signal is weak compared to water and other molecules in tissue, making it hard to isolate. Raman spectroscopy, which uses laser light to identify molecules by how they scatter photons, has shown promising accuracy in animal studies, with a correlation coefficient of 0.91 against standard meter readings in mice.
Bioimpedance spectroscopy measures how a small electrical signal passes through tissue, since glucose concentration affects the electrical properties of cells. Prototype sensors have achieved resolution as fine as 4 mg/dL, but accuracy drops significantly outside of controlled lab settings, where factors like skin hydration, temperature, and movement introduce noise. Reverse iontophoresis uses a gentle electrical current to draw glucose-containing fluid through the skin for measurement on the surface.
None of these technologies have replaced finger pricks or CGMs in clinical practice. The core challenge is the same across all of them: glucose exists in very low concentrations compared to the other substances in your body, making it difficult to measure precisely through intact skin.