Compare
Home A1c tests: how accurate are they?
An at-home A1c kit uses a drop of blood from a fingerstick. It can show roughly where you are between lab tests. Here is how far a kit result can sit from a lab result, and what a kit cannot do.
The short answer
- Not for diagnosis. The ADA says screening and diagnosis should use devices the FDA approved for diagnosis, in certified labs with trained staff (ADA Standards 2026, sec. 2). A home kit cannot confirm or rule out diabetes.
- Can be close, but the margin is wider. In one study of a home kit, 93% of results by lay users were within 13.5% of the lab value (Chang 2010). US lab proficiency testing allows 8% (NGSP news).
- Check the certification. NGSP certifies A1c methods for one year at a time, for specific reagents and instruments (NGSP criteria). Its public list shows which methods are current (NGSP list).
What "accurate" means in numbers
Accuracy rules for A1c are written as a percent of the true value, not as A1c points. Here is what each yardstick allows around a lab result:
| Lab A1c | Within 5% | Within 8% | Within 13.5% |
|---|---|---|---|
| 6.5% | 6.2% to 6.8% | 6.0% to 7.0% | 5.6% to 7.4% |
| 7.0% | 6.7% to 7.4% | 6.4% to 7.6% | 6.1% to 7.9% |
| 8.0% | 7.6% to 8.4% | 7.4% to 8.6% | 6.9% to 9.1% |
| 9.0% | 8.6% to 9.5% | 8.3% to 9.7% | 7.8% to 10.2% |
- Within 5%: NGSP certification for manufacturers (36 of 40 samples).
- Within 8%: US lab proficiency testing since 2025.
- Within 13.5%: Counted as accurate in the Chang 2010 home-kit study.
Simple arithmetic: lab value × (1 ± percent), rounded to one decimal. Sources: NGSP criteria (NGSP criteria), NGSP on proficiency testing (NGSP news), Chang 2010 (Chang 2010).
So with a true A1c of 7.0%, a home result anywhere from 6.1% to 7.9% would have counted as accurate in that study. That spread crosses several key values, which is why a home result is a guide, not a verdict.
What the studies found
- A home kit used by lay people. 110 adults (93 with diabetes) tested themselves with an over-the-counter kit, and results were compared with a lab analyzer. 165 of 177 self-test results (93.2%) were within 13.5% of the lab value. The first-time failure rate was 39.6% without the instruction video and 11.3% with it (Chang 2010).
- Professional point-of-care analyzers. A Dutch lab tested 8 clinic analyzers. Two manufacturers withdrew after early results, and only 2 of the 8 met the accepted performance criteria (Lenters-Westra 2010).
- A well-performing clinic analyzer. In a 2019 US study, fingerstick results run by clinic staff on one point-of-care analyzer differed from the lab by 0.2 percentage points on average (Nathan 2019).
Devices differ, and so do lots of the same device: the Dutch study found significant differences between reagent lots for every point-of-care instrument it tested (Lenters-Westra 2010).
How to check a kit
- Look for the kit's method on the NGSP list of certified methods, and check that the certificate is current. Certificates expire after one year (NGSP list).
- Follow the instructions exactly. In the home-kit study, many first-time users failed the test without the instruction video (Chang 2010).
- Compare one home result with a lab result taken around the same time, so you know how your kit reads for you.
When a home result and a lab result disagree
- Trust the lab result for decisions, and tell your care team about the gap. A lab test is the standard for diagnosis (ADA Standards 2026, sec. 2).
- Two lab tests on the same sample can also differ a little: NIDDK's example is a 6.8% that could read 6.4% to 7.2% on a repeat (NIDDK).
- Some conditions affect A1c whatever the device. See when A1c can mislead.
Sources
- American Diabetes Association Professional Practice Committee. 2. Diagnosis and Classification of Diabetes: Standards of Care in Diabetes 2026. Diabetes Care. 2026;49(Suppl 1). doi.org/10.2337/dc26-S002
- Chang A, Frank J, Knaebel J, Fullam J, Pardo S, Simmons DA. Evaluation of an over-the-counter glycated hemoglobin (A1C) test kit. J Diabetes Sci Technol. 2010;4(6):1495-1503. doi.org/10.1177/193229681000400625
- Lenters-Westra E, Slingerland RJ. Six of eight hemoglobin A1c point-of-care instruments do not meet the general accepted analytical performance criteria. Clin Chem. 2010;56(1):44-52. doi.org/10.1373/clinchem.2009.130641
- Nathan DM, Griffin A, Perez FM, Basque E, Do L, Steiner B. Accuracy of a point-of-care hemoglobin A1c assay. J Diabetes Sci Technol. 2019;13(6):1149-1153. doi.org/10.1177/1932296819836101
- NGSP. Summary of certification criteria (manufacturers: 36 of 40 results within 5% of a Secondary Reference Laboratory; certificate valid 1 year, specific to reagents and instruments). ngsp.org/critsumm.asp
- NGSP. List of NGSP certified methods (updated monthly; certificates expire after 1 year). ngsp.org/certified.asp
- NGSP. CMS rule: from January 1, 2025, US HbA1c proficiency testing uses acceptable limits of plus or minus 8% (CAP GH-5: plus or minus 6%). ngsp.org/news.asp
- National Institute of Diabetes and Digestive and Kidney Diseases (NIH). The A1C Test & Diabetes. www.niddk.nih.gov/health-information/diagnostic-tests/a1c-test
Written and maintained by the controla1clevels.com editorial team. We do not have a medical reviewer; every claim links to its primary source above. Last reviewed: 2026-10-26. Spotted an error? See About.