Are at-home blood tests accurate?

What decides whether an at-home result is accurate

An at-home blood test has two halves. You collect the sample, and a laboratory analyses it. Accuracy depends on a short chain of links, and a weak link anywhere can skew a result:

• The lab. Under CLIA, the federal programme run by CMS, labs that test human samples must meet quality standards designed to make results accurate, reliable and timely.
• The collection device. Devices that lance the skin to draw blood are regulated by the FDA as medical devices, and a 510(k) clearance covers the specific indications for use stated in it.
• Sample volume. The lab needs enough blood for every test on the panel. In a 2025 study of 120 people, 94% of self-collections produced a sample that could be analysed for at least one test.
• Sample quality. Haemolysis, the breakdown of red blood cells in the sample, is the most common reason labs reject samples, and it distorts some markers far more than others.
• Transport and time. Blood keeps changing after it leaves the body. Researchers in the same 2025 study judged home sampling reliable only for the markers that stayed stable until the lab processed them.

Capillary versus venous blood: what the studies show

A standard blood test draws venous blood from a vein in your arm. At-home tests use capillary blood from tiny vessels just under the skin. The question is whether the two give the same answer. For several common markers, recent studies say they largely do.

In a 2026 study in the Journal of Diabetes Investigation, 105 people collected their own capillary sample with a self-collection device and sent it by standard mail. Their HbA1c results differed from couriered venous samples by less than 0.5% on average, with a concordance correlation of 0.999, where 1.0 is perfect agreement. Total, LDL and HDL cholesterol showed average differences of 3.0% to 5.7% and concordance above 0.95.

A 2025 study from the Netherlands Cancer Institute, published in Clinical Chemistry and Laboratory Medicine, asked 120 people (21 healthy volunteers and 99 patients) to self-collect capillary blood after a venous draw. Total, HDL and LDL cholesterol, triglycerides, ALT, CRP and total protein all met the tightest of three accuracy standards, which are based on how much each marker naturally varies. Albumin met the minimum standard.

Where agreement is weaker

Not every marker passed. In the 2025 study, glucose and AST missed even the minimum standard. AST passed only once the authors applied a correction to it.

AST is also unusually sensitive to haemolysis. In a laboratory study that broke red cells on purpose, AST and LD (lactate dehydrogenase) were affected at haemolysis levels almost too low to see. Potassium and bilirubin shifted meaningfully only with moderate haemolysis. ALT, cholesterol and GGT held steady even when haemolysis was severe. The authors recommended that labs measure haemolysis directly and request a new sample for markers it affects.

Glucose is the other weak spot. If a glucose or AST result from a home sample looks out of line with the rest of your panel, a doctor can repeat it with a venous draw.

Self-collection devices versus finger-lance blood spots

Traditional blood spot kits ask you to lance a fingertip and drip blood onto filter paper, where it dries. Self-collection devices like those in the studies above draw liquid capillary blood into a tube, which the lab can process into serum and run on the same analysers it uses for venous samples.

Dried blood spots do well for HbA1c. A 2014 meta-analysis of 16 studies in BMC Clinical Pathology found a near one-to-one relationship between spot and venous HbA1c. Cholesterol was less reliable. The pooled line for total cholesterol had a slope of 0.68 rather than 1, so spot results drifted away from venous results as cholesterol rose. There was only one study for HDL and none for LDL, and the authors called for standard methods for collection, transport, storage and analysis.

A US study within the National Health and Nutrition Examination Survey (NHANES) reached a similar verdict. Blood spots collected in participants' homes by survey staff correlated well with venous HbA1c (r = 0.90) but poorly with total cholesterol (r = 0.65 to 0.67) and HDL cholesterol (r = 0.64 to 0.66), and the authors judged spots not comparable for those two markers. The venous and spot samples were taken one to three weeks apart, but the authors pointed mainly to the lab methods used for spot lipids.

The device alone does not settle accuracy. For cholesterol, though, the studies above with the closest agreement used liquid capillary samples run on standard lab analysers.

How to check an at-home test provider

You can check the two most important links in the chain yourself, using free federal databases:

• Find the lab's CLIA number, a 10-character code. With some exceptions and state exemptions, CMS says every lab that tests human samples for health purposes needs a current CLIA certificate, including labs that never bill Medicare. CLIA covers about 320,000 laboratory entities.
• Look the number up in the QCOR CLIA Laboratory Lookup, linked from the CMS CLIA webpage.
• Check the certificate type. A Certificate of Waiver covers only simple tests the FDA rates as waived, such as certain glucose and cholesterol tests. Labs running moderate- and high-complexity tests hold a Certificate of Compliance, issued after an on-site survey and resurveyed every two years, or a Certificate of Accreditation from one of 7 CMS-approved accreditation organisations, which inspect every two years.
• Those labs must also take part in proficiency testing for certain tests: an approved programme sends test samples 3 times a year, the lab tests them like patient samples, and the programme grades the results.
• Check the collection device. Search its 510(k) number in the FDA's 510(k) Premarket Notification database. A clearance means the FDA found the device substantially equivalent to one already legally on the market, for the indications for use stated in the clearance. Read the letter and summary attached to the record, including any limitations on how the device may be used.
• Ask the provider who reviews your results, and what happens if your sample arrives too small or haemolysed.

When to speak to a doctor

Treat a single home result as one data point, not a diagnosis. If a result is unexpected, far outside the reference range on your report or does not match how you feel, speak to a doctor, who may repeat the test with a venous draw. Never start, stop or change a medication or supplement because of an at-home result without talking to your doctor first.