Four blood tests are widely used to diagnose diabetes: the hemoglobin A1c (HbA1c), the fasting plasma glucose, the oral glucose tolerance test, and the random plasma glucose. Each measures blood sugar differently, and the one your doctor orders depends on what they are looking for and your individual circumstances. Beyond diagnosis, a broader panel of blood tests can reveal what type of diabetes you have, how well treatment is working, whether complications like kidney disease are developing, and whether a dangerous crisis like diabetic ketoacidosis is underway.
The Four Core Diagnostic Tests
The American Diabetes Association recognizes four biochemical routes to a diabetes diagnosis: a fasting plasma glucose of 126 mg/dL or higher, a random plasma glucose of 200 mg/dL or higher with classic symptoms, a two-hour oral glucose tolerance test result of 200 mg/dL or higher, or an HbA1c of 6.5% or higher.1PubMed Central. Diagnostic utility of fasting versus non-fasting glucose: contextualising testing strategies-a narrative review In practice, if one test comes back above the threshold, a second confirmatory test is usually done before a formal diagnosis is made, unless you already have unmistakable symptoms like excessive thirst, frequent urination, and unexplained weight loss paired with a high random glucose.
HbA1c
HbA1c reflects your average blood sugar over the past two to three months. It works by measuring how much glucose has attached to hemoglobin, the oxygen-carrying protein in red blood cells. Because red blood cells live roughly 120 days, the test captures a rolling average rather than a single snapshot. You do not need to fast for this test, which makes it convenient for both patients and clinics.
An HbA1c below 5.7% is considered normal. The prediabetes range sits between 5.7% and 6.4%, and 6.5% or above meets the diagnostic threshold for diabetes. Within that prediabetes band, however, risk is not evenly distributed. A systematic review of HbA1c thresholds found that the 5.7–6.4% window identifies a large number of adults, but many in the lower end of that range have low short-term risk of progressing to full-blown diabetes. People in the 6.0–6.4% range, and especially 6.2–6.4%, face much higher near-term risk.2PubMed Central. Glycated Hemoglobin and Prediabetes: A Systematic Review of HbA1c Thresholds for Type 2 Diabetes Prevention So a result of 5.8% and a result of 6.3% both technically fall under “prediabetes,” but they carry very different implications for what happens next.
For people already living with diabetes, HbA1c is also the standard monitoring tool. Most guidelines recommend keeping it below 7% for the average adult with diabetes, though your doctor may set a tighter or looser target depending on your age, how long you have had diabetes, and other health conditions.
Fasting Plasma Glucose
A fasting plasma glucose test requires you to go without food or caloric drinks for 8 to 12 hours, usually overnight. It measures how well your body manages blood sugar when it has not had any incoming fuel. In a healthy person, fasting glucose typically sits around 80–90 mg/dL, with the liver producing just enough glucose to match what the body is burning.1PubMed Central. Diagnostic utility of fasting versus non-fasting glucose: contextualising testing strategies-a narrative review
A fasting glucose below 100 mg/dL is normal. Between 100 and 125 mg/dL is classified as impaired fasting glucose, which is one form of prediabetes. At 126 mg/dL or above, diabetes is diagnosed. The test is straightforward and inexpensive, but it has a practical drawback: the fasting requirement means you need to plan ahead. If you ate breakfast before arriving at the lab, the result is not valid for diagnostic purposes.
Oral Glucose Tolerance Test
The oral glucose tolerance test, or OGTT, pushes harder than a fasting glucose test. After fasting overnight, you drink a sugary solution containing 75 grams of glucose and then have your blood drawn two hours later. It specifically tests how effectively your body clears a large sugar load. A two-hour result below 140 mg/dL is normal. Between 140 and 199 mg/dL indicates impaired glucose tolerance, and 200 mg/dL or higher means diabetes.3PubMed. Glucose tolerance test – non-pregnant
The OGTT can catch problems that fasting glucose alone misses, because some people maintain a normal fasting level but cannot handle the spike after eating. This test is less commonly used in routine screening for adults because it requires two hours of sitting in a clinic, but it remains the gold standard for diagnosing gestational diabetes and is sometimes used when other tests give borderline or conflicting results.
When Test Results Disagree
Here is something that catches people off guard: it is entirely possible for your HbA1c and fasting glucose to point in different directions. One might suggest prediabetes while the other looks normal, or vice versa. This discordance is actually well-documented, particularly in people who experience reactive hypoglycemia, where blood sugar drops sharply after meals.4INTERNATIONAL JOURNAL OF SCIENTIFIC RESEARCH. DISCORDANCE BETWEEN HBA1C AND FASTING PLASMA GLUCOSE IN REACTIVE HYPOGLYCEMIA: A CASE SERIES In those cases, the low post-meal dips can average out the high spikes, leaving HbA1c looking reassuringly normal even though glucose handling is clearly abnormal.
The tests are also measuring fundamentally different things. Fasting glucose captures one moment in time. HbA1c captures a weighted average over months. The OGTT captures your response to a specific challenge. A person whose blood sugar spikes dramatically after meals but returns to normal by morning could have a normal fasting glucose but an elevated HbA1c or OGTT. When results conflict, most clinicians repeat the abnormal test or add a different one to get a clearer picture.
Why HbA1c Can Be Wrong
HbA1c is a reliable test for most people, but it has real blind spots. Anything that affects hemoglobin or red blood cell lifespan can throw the number off. Iron deficiency anemia, for instance, tends to falsely raise HbA1c. In one study comparing different types of anemia in people without diabetes, the average HbA1c in iron-deficient individuals was 5.75% compared to 5.32% in healthy controls. After the iron deficiency was treated, HbA1c dropped significantly to 5.44%.5PubMed Central. The effect of different types of anemia on HbA1c levels in non-diabetics Sickle cell anemia also pushed HbA1c upward in the same study, with a mean of 5.83%.
Hemoglobin variants cause a separate problem. Certain inherited hemoglobin types can directly interfere with the laboratory machines that measure HbA1c. While most modern analyzers handle common variants like HbS, HbC, HbD, and HbE without trouble, elevated fetal hemoglobin (HbF) causes significant errors across multiple testing platforms.6PubMed. Interference of hemoglobin variants with HbA1c measurements by six commonly used HbA1c methods The broader concern is that genetic hemoglobin variants can interfere with several different types of HbA1c assay chemistry.7PubMed. Interference of hemoglobin variants in HbA(1c) quantification If you carry a hemoglobin trait and your HbA1c seems inconsistent with your day-to-day glucose readings, this is a likely explanation. In these situations, doctors may rely more heavily on fasting glucose or the OGTT.
Glycated Albumin and Fructosamine
When HbA1c is unreliable due to anemia, hemoglobin variants, or conditions that shorten red blood cell lifespan (such as chronic kidney disease on dialysis), doctors can turn to alternative glycemic markers. Glycated albumin and fructosamine both measure how much sugar has attached to proteins in the blood, but they reflect a shorter window of roughly two to three weeks instead of two to three months.
Glycated albumin in particular tracks more closely with actual blood sugar levels than fructosamine does. Research comparing the two found that glycated albumin correlated well with HbA1c and with preceding mean blood glucose, while fructosamine did not correlate as reliably with mean glucose.8PubMed. Blood glycated haemoglobin, serum fructosamine, serum glycated albumin and serum glycated total protein as measures of glycaemia in diabetes mellitus Glycated albumin also responds faster to treatment changes, dropping more rapidly than HbA1c when blood sugar control improves, which makes it useful for tracking whether a new medication or insulin regimen is working.9PubMed. Comparison of glycated albumin (GA) and glycated hemoglobin (HbA1c) in type 2 diabetic patients: usefulness of GA for evaluation of short-term changes in glycemic control
C-Peptide and Autoantibody Tests for Typing Diabetes
A blood sugar test can tell you that your glucose is too high, but it does not tell you why. That distinction matters because the treatment for type 1 diabetes, type 2 diabetes, and rarer forms like monogenic diabetes can be very different. Two additional blood tests help sort this out.
C-peptide is a byproduct of insulin production. When the pancreas makes insulin, it first produces a precursor molecule called proinsulin, which is then cleaved into insulin and C-peptide in equal amounts. Measuring C-peptide tells your doctor how much insulin your pancreas is still making.10PubMed Central. A Practical Review of C-Peptide Testing in Diabetes A very low C-peptide suggests the insulin-producing beta cells are mostly gone, pointing toward type 1 diabetes. A normal or high C-peptide alongside high blood sugar suggests the body is making plenty of insulin but not responding to it properly, which is the hallmark of type 2 diabetes. C-peptide is preferred over directly measuring insulin because insulin is cleared from the blood much faster, with a half-life of just 3 to 5 minutes compared to 20 to 30 minutes for C-peptide, giving a more stable measurement.11PubMed Central. C-peptide in Precision Diabetes Care and Beyond: A Comprehensive Review
Autoantibody testing goes a step further. Type 1 diabetes is an autoimmune condition in which the immune system attacks the pancreatic beta cells. The attack leaves detectable antibodies in the blood. Screening for these diabetes-related autoantibodies can identify people in early stages of type 1 diabetes before symptoms even appear, sometimes years in advance.12PubMed Central. Recommendations for Screening and Monitoring the Stages of Type 1 Diabetes in the Immune Therapy Era This has become increasingly relevant now that stage-specific therapies exist to delay the onset of clinical diabetes in high-risk individuals. Autoantibody testing is also used when an adult is diagnosed with diabetes and the clinical picture is ambiguous, helping distinguish late-onset type 1 from type 2.
Gestational Diabetes Screening
Pregnant women are typically screened for gestational diabetes between 24 and 28 weeks of pregnancy, though those with risk factors may be tested earlier. The most common approach in the United States is a two-step process: first a one-hour glucose challenge test (drinking 50 grams of glucose and checking blood sugar an hour later), and if that is elevated, a follow-up three-hour OGTT with 100 grams of glucose.
Internationally, a one-step method using the criteria from the International Association of the Diabetes and Pregnancy Study Groups (IADPSG) is increasingly used. This involves a single 75-gram OGTT with lower glucose thresholds, which identifies more women as having gestational diabetes. A comparison study found that prevalence was about 7% when using the older Carpenter-Coustan criteria but roughly 23% when using the IADPSG criteria in the same hospital setting.13PubMed Central. A Comparison of Pregnancy Outcomes Using Two Diagnostic Criteria for Gestational Diabetes Mellitus-Carpenter Coustan Criteria and International Association of the Diabetes and Pregnancy Study Groups (IADPSG) Criteria Which method your provider uses matters because it directly affects whether you receive a diagnosis and, consequently, treatment. HbA1c is generally not recommended as the primary diagnostic tool for gestational diabetes because the physiological changes of pregnancy, including increased red blood cell turnover, can make it less accurate.
Ketone Testing in Diabetic Emergencies
When blood sugar runs dangerously high and insulin is severely deficient, the body starts breaking down fat for fuel and produces ketones as a byproduct. This can spiral into diabetic ketoacidosis (DKA), a life-threatening emergency most associated with type 1 diabetes but also seen in some people with type 2. The key blood test here is beta-hydroxybutyrate (BOHB), the predominant ketone body in the bloodstream during DKA.
In emergency settings, a BOHB level of about 5.3 mmol/L predicted DKA with roughly 91% overall accuracy in a study of pediatric emergency department patients.14PubMed Central. Plasma Beta-Hydroxybutyrate for the Diagnosis of Diabetic Ketoacidosis in the Emergency Department Point-of-care ketone meters that measure BOHB from a fingerstick are widely available and increasingly used by people with type 1 diabetes at home to catch early ketone elevations before they progress to a crisis. The older urine ketone strips, while still in use, are less precise and lag behind blood levels.
Kidney Function Tests in Diabetes
Diabetes is the leading cause of chronic kidney disease worldwide, and early detection of kidney damage is one of the most important reasons for regular lab work in anyone with diabetes. The standard screening involves two tests: a urine albumin-to-creatinine ratio, which checks for small amounts of protein leaking into the urine, and a serum creatinine to estimate how well the kidneys are filtering blood (the estimated glomerular filtration rate, or eGFR).
A newer blood marker, cystatin C, is gaining ground as a more sensitive early indicator of declining kidney function in people with diabetes. Cystatin C-based equations tend to detect kidney function decline earlier than creatinine-based ones, particularly in the early stages of diabetic kidney disease.15PubMed Central. Cystatin C-based equations: Enhancing accuracy in kidney function tests for type 2 diabetes One study at a teaching hospital found that cystatin C flagged reduced kidney filtration in 23% of diabetes patients who did not yet have protein in their urine, compared to only 13% flagged by creatinine-based estimates.16PubMed Central. Diagnostic Performance of Cystatin C in the Early Detection of Diabetic Kidney Disease at the University of Nigeria Teaching Hospital, Ituku-Ozalla Creatinine is affected by muscle mass, diet, and other factors, so cystatin C can be especially useful in people where creatinine might be misleading, like those with very low or high muscle mass.
Home Glucose Meters and Continuous Glucose Monitors
For day-to-day management, most people with diabetes rely on either fingerstick glucose meters or continuous glucose monitors (CGMs). These are monitoring tools, not diagnostic ones. Point-of-care glucose meters are explicitly not intended for diagnosing diabetes because they lack the standardization and traceability required for a clinical diagnosis.17PubMed Central. Comparative Performance Analysis of Three Point-of-Care Glucose Testing Devices against a Central Laboratory Reference Method If your home meter reads high, it is a reason to see your doctor for proper lab testing, not a diagnosis in itself.
Fingerstick meters also measure capillary blood from your fingertip, while lab tests use venous blood drawn from a vein. These two sources do not always give the same result, and converting between them is not straightforward.18PubMed. Comparability of venous and capillary glucose measurements in blood The difference tends to be small for most practical purposes, but it is large enough to cause a misclassification in glucose tolerance status around diagnostic thresholds.
CGMs, worn on the skin and measuring interstitial glucose every few minutes, generate a metric called the glucose management indicator (GMI), which estimates what your HbA1c might be based on your average sensor glucose. But the relationship between CGM-derived GMI and lab-measured HbA1c is not as tight as many people assume. In a study of over 500 people with diabetes, only about 19% had GMI and HbA1c values that matched exactly, and in 28% the two differed by more than half a percentage point.19PubMed Central. A View Beyond HbA1c: Role of Continuous Glucose Monitoring This mismatch is not a flaw in the CGM; it reflects the biological reality that people glycate hemoglobin at different rates for the same average glucose. Your HbA1c might always run higher or lower than your CGM data would predict.
Monogenic Diabetes and Genetic Testing
Not all diabetes fits neatly into the type 1 or type 2 categories. Monogenic diabetes, sometimes called MODY (maturity-onset diabetes of the young), is caused by a single gene mutation and accounts for a small but significant fraction of diabetes cases. It is commonly misdiagnosed as type 1 or type 2 because the blood sugar tests look the same. What sets it apart is the clinical picture: onset in young adults, a strong family history of diabetes spanning multiple generations, and sometimes very mild hyperglycemia that does not require insulin.
Standard blood tests like HbA1c and fasting glucose cannot distinguish monogenic diabetes from other types. Biomarkers and clinical scoring systems can help flag candidates, but a definitive diagnosis requires genetic testing.20PubMed Central. Monogenic diabetes: An evidence-based clinical approach Getting the right diagnosis matters because treatment often changes dramatically. Some forms of MODY respond well to low-dose sulfonylurea pills and do not need insulin at all, while others require no treatment beyond monitoring. In a real-world study of young Thai patients with atypical diabetes who underwent genetic testing, about 12% were found to carry a disease-causing variant, and the genetic diagnosis improved clinical management for both the patients and their family members.21PubMed Central. Identification of maturity-onset diabetes of the young through targeted next-generation sequencing in Thai patients with atypical diabetes in real-world practice
Lab Standardization and Why It Matters
One thing most people never think about is whether the lab running their HbA1c test is producing comparable results to other labs. For decades, this was a real problem. Different instruments using different chemical methods could give meaningfully different HbA1c values for the same blood sample. The National Glycohemoglobin Standardization Program (NGSP), working alongside an international counterpart, has spent over 20 years harmonizing HbA1c methods so that a result of 7.0% means the same thing regardless of which lab ran it.22PubMed Central. The National Glycohemoglobin Standardization Program: Over 20 Years of Improving Hemoglobin A(1c) Measurement New HbA1c analyzers must be certified through this system before they can be used clinically.23PubMed. Evaluating new HbA1c methods for adoption by the IFCC and NGSP reference networks using international quality targets For glucose testing, similar standardization exists through a separate traceability chain. The practical takeaway is that lab results are far more reliable and comparable today than they were even 20 years ago, though no standardization program eliminates every source of error, particularly the biological confounders like hemoglobin variants discussed earlier.