Can a paternity test be wrong?

Rarely, and most often because of a sample problem, such as a mix-up or contamination, or because the man tested is closely related to the real father. Identical twins are the one case a standard test can't settle.

Updated November 14, 2026

A paternity test compares the child's DNA with the possible father's at a set of markers. Accredited labs report results that are close to all or nothing: Labcorp says its results usually show a probability of paternity of either 0.00% or 99.99% or greater. Mistakes come less from the science than from what happens around it: whose sample went into the tube, how it was handled, and whether the right men were tested.

What can make a paternity test wrong

  • Mixed-up or contaminated samples. Geneticists at The Tech Interactive name human error, such as swapped samples, and contamination as the main ways a test goes wrong. AABB-accredited labs keep a documented chain of custody from the start of a case to the report to guard against this.
  • No identity check. With a home kit, nobody confirms who gave each swab. Cleveland Clinic notes that at-home results may not be as accurate and aren't admissible in court.
  • A related possible father. If the real father is the tested man's brother, father or son, the tested man shares many markers with the child. A standard test may not tell them apart unless both men are tested.
  • A mutation. DNA occasionally changes between generations, so one marker can mismatch even when the man is the father. Labs don't exclude a man over a single mismatch. The New York City medical examiner's protocol, for example, requires two or more.
  • Rare biology. Chimerism, where one person carries two sets of DNA, can in rare cases mislead a test.

Can a paternity test be wrong if the possible fathers are brothers?

It can be, if only one brother is tested and the lab isn't told. Brothers who aren't identical twins have different DNA, but they share more markers than unrelated men, so one brother can appear to match when the other is the father. DNA Diagnostics Center (DDC) recommends three steps:

  1. Tell the lab up front that the possible fathers are brothers.
  2. Test both brothers with the child.
  3. Include the child's mother.

The lab can then test extra markers to tell the brothers apart. Each man must agree to be tested. The same applies when the possible fathers are a father and his son.

Can a DNA test tell identical twins apart?

Not a standard paternity test. Identical twins have the same DNA at the markers these tests use, so if the possible fathers are identical twins, both will match. DDC calls telling them apart with standard testing practically impossible. The law reflects this. In Texas, a court can order testing of an identical brother and, if both brothers match, decide which is the father on other evidence.

Research labs have gone further. In a 2014 study, scientists sequenced twin brothers' DNA far more deeply than any paternity test does and found five rare changes that the father and child shared but his twin did not. That is research-grade work, not a routine test.

What is the most accurate paternity test?

The strongest result comes from a test that:

  • Includes the mother. Labcorp prefers to test her when possible. Her sample shows which half of the child's DNA came from her, which leaves the father's half clear.
  • Includes every possible father. This matters most when the men are related.
  • Checks enough markers. Cleveland Clinic says the industry standard is 16 and that more markers give more accurate results.
  • Uses an AABB-accredited lab and a witnessed collection. AABB says its accredited labs follow validated procedures to keep a valid chain of custody from the start of a case to the report.

Before birth, Cleveland Clinic describes the noninvasive prenatal paternity test, which uses a blood sample from the mother, as 99.9% accurate. See paternity testing while pregnant.

What to do if you think your result is wrong

  1. Call the lab. AABB requires accredited labs to have a complaint process. If that doesn't resolve it, you can contact AABB.
  2. Retest with a legal test. Use a witnessed collection at an AABB-accredited lab, and include the mother and every possible father.
  3. In a court or child support case, act quickly. States set deadlines for written objections to results. When a child support agency ordered the test, federal law lets a party who contests it get more testing by paying in advance.

Our guide to reading paternity test results shows how to spot an exclusion or a single mismatch on your report.

If you plan to retest with a witnessed, chain-of-custody collection, start with our legal paternity test.

Questions

Can a 99.99% paternity test be wrong?

Only through a sample or handling error, or when the real father is a close relative or identical twin of the man tested. The figure compares the man with a random unrelated man, not with his relatives.

Can a 0% paternity test be wrong?

A 0% result means the man lacks the father's DNA at more than one marker. Unless the samples were mixed up, he is not the biological father.

Can a home paternity test be wrong?

Yes, in the same ways as any test. With a home kit, nobody checks who gave each swab, so a mix-up or swapped sample can go unnoticed. Home results also can't be used in court.

Sources

Related guides