[Expert Explanation] Can You Trust a 99.9% Paternity Positive Probability? — Why Positive Results Are More Reliable Than Negative Results in DNA Testing

Rewritten on: January 9, 2026

This article explains, based on scientific evidence and international standards, the meaning and reliability of a paternity positive probability of 99.9% or higher, the difference from a negative result (0%), and genetic exceptions such as chimerism.

In DNA testing, you often see the expression "paternity positive probability of 99.9%." So how reliable is this number really? In conclusion, a positive result of 99.9% or higher is extremely strong evidence that "the parties are almost certainly parent and child." On the other hand, while negative results (0%) are also usually accurate, they can, in rare cases confirmed worldwide, be overturned by genetic exceptions (such as chimerism).

Among those considering a DNA parentage test, many wonder, "Can I really trust 99.9%?" or "Why isn't it 100%?" Others feel anxious when a negative result comes back, wondering "Are we really not parent and child?" This article clearly explains the mechanisms behind positive and negative results, the difference in their reliability, and why exceptions occur, based on scientific evidence and international standards.

What Does a 99.9% Paternity Positive Probability Mean?

A paternity positive probability is "an index showing how strongly a DNA match supports the paternity relationship, rather than being the result of chance." In DNA testing, more than 20 STR (short tandem repeat) sites are compared, and the paternity probability is calculated from the degree of matching. [ref:1]

STR stands for "Short Tandem Repeat," referring to short repeating segments of base sequences found throughout the human genome. Because the number of repeats differs between individuals, STRs are used as extremely useful genetic markers for personal identification and parentage testing. Current standard DNA testing analyzes 20 or more STR markers simultaneously, achieving extremely high accuracy in determinations. [ref:2]

The general interpretation is as follows:

1. 99.9% or higher
   This cannot be explained by chance and means it is almost certain that the man is the father. The probability that an unrelated man would coincidentally match this many STR markers is less than 1 in several hundred billion, making it a level at which "parent and child" can effectively be stated with certainty.
2. 0%
   This means that, based on the genetic data, a parent-child relationship is statistically excluded. This result occurs when clear mismatches are confirmed at multiple STR markers, and alleles (genotypes) that could not have been inherited from the candidate father are detected.

International standards such as AABB (Association for the Advancement of Blood & Biotherapies) require 99% or higher for legal testing. In practice, results of 99.9% to 99.99% are common, and these are treated as "effectively conclusive." The reason the paternity positive probability never reaches 100% is based on statistical principles. In probability theory, it is theoretically impossible to declare that "a hypothesis is correct" with 100% certainty, no matter how much evidence accumulates. However, a figure of 99.99% can reasonably be regarded as "certain" for everyday decision-making purposes. [ref:3] [ref:4]

Why Are Positive Results More Reliable Than Negative Results?

■ Positive results are supported by "numerous matches"
When many STR markers match without contradiction, the probability of that match occurring by chance is said to be less than 1 in several hundred billion. This is because there is extremely strong statistical support in the form of the candidate father's alleles matching the child's alleles across all 20 or more independent genetic markers. As a result, once a positive result is established, there is almost no possibility the result will change. [ref:2]

■ Negative results are determined by "a small number of mismatches"
In DNA testing, a result of 0% occurs when several decisive mismatches are found. Typically, a parent-child relationship is denied when mismatches are confirmed at three or more STR markers. However, these mismatches can sometimes be caused by the following:

• Chimerism (discussed below) — an extremely rare genetic condition in which multiple sets of DNA coexist within one body
• Sample contamination — cases where a third party's DNA becomes mixed in during collection, transport, or analysis
• Special cases related to fertility treatment (IVF) — extremely rare reports of embryo mix-ups
• Documentation mix-ups — human errors such as mislabeled samples or errors in report entries

In this way, the key point is that negative results carry a slight residual possibility of misjudgment due to exceptions. While positive results are backed by the overwhelming statistical evidence of "numerous matches," negative results are based on "a small number of mismatches," so there are cases where it is necessary to carefully verify whether the cause of the mismatch is truly the absence of a biological relationship. [ref:6]

Preconditions for a Negative (0%) Result to Hold

For a negative result (0%) to be completely correct, the following conditions must be met:

1. The collected sample must accurately reflect the individual's genetic information
   Cells or blood samples collected from the oral mucosa must reliably represent that individual's DNA. In the extremely rare case of chimerism, DNA type may differ depending on the site of collection.
2. There must be no contamination, mix-up, or storage-related problems
   Throughout the entire process from sample collection to completion of analysis, there must be no mixing in of a third party's DNA, mislabeling, or degradation due to improper storage conditions. Reliable testing laboratories strictly maintain a chain of custody (evidence preservation management system).
3. There must be no genetic exceptions such as chimerism or mosaicism
   It is assumed that the chimerism or mosaicism phenomena discussed below do not exist in the test subject. These genetic phenomena are extremely rare, but if present, they can have a significant impact on the test result.

Since these preconditions are met in most cases, negative results are also generally very reliable. However, this is precisely the source of the reliability gap where "positive > negative." A positive result's conclusion is unlikely to change even if the above preconditions are somewhat shaky, whereas for a negative result, there is a slight risk of misjudgment if even one of these preconditions fails to hold.

What Is "Chimerism," Which Can Confuse DNA Testing?

Chimerism refers to a condition in which different DNA coexists within a single body. Normally, since a human develops from a single fertilized egg, all cells throughout the body carry identical DNA. However, in the case of chimerism, during the fetal period, one embryo of a pair of fraternal (dizygotic) twins may be absorbed into or fused with the other, resulting in different parts of the body carrying different genetic information. [ref:7]

A well-known case is the "Karen Keegan case" reported in the United States. She drew attention when a kidney transplant compatibility test using her blood DNA unexpectedly denied a maternal relationship with her own biological children. A subsequent detailed investigation revealed that she was a "tetragametic chimera," carrying different DNA in her blood and in her germline (ovarian tissue). [ref:5]

In such a condition, because DNA differs by body site, a mismatch can occur even between an actual parent and child, resulting in a 0% (false-negative) result. Specifically, differing DNA types have been reported in the following tissues:

• Cases where blood and saliva show different DNA types
• Cases where germ cells (sperm/eggs) and somatic cells carry different DNA types
• Cases where different DNA types are detected between the left and right sides of the body, such as in skin or hair

That said, chimerism is an extremely rare phenomenon worldwide. Only a few dozen clinically detected cases of chimerism are known worldwide, and it is essentially never encountered in routine testing. If there is ever doubt about a test result, it is possible to verify the possibility of chimerism by retesting using a different sample (for example, blood or hair roots instead of oral mucosa).

There is also a concept similar to chimerism called "mosaicism." Mosaicism is a phenomenon in which, after development from a single fertilized egg, a mutation occurs in some cells, resulting in genetically distinct cell populations within the same individual. Chimerism is distinguished by originating "from two different fertilized eggs," while mosaicism "originates from a single fertilized egg but develops a mutation along the way." [ref:8]

Differences Between Positive and Negative Results (Comparison Table)

The table below summarizes the main differences between positive and negative paternity results.

• Legal evaluation (positive): Treated as "effectively conclusive" in many countries and accepted as strong evidence in court.
• Legal evaluation (negative): Usually accepted as a denial, but retesting may be requested since there is room to consider exceptional cases.
• Note (positive): Reliability depends heavily on the testing lab's quality control system.
• Note (negative): If an exception is suspected, retesting using a different sample is necessary.

Testing Lab Quality That Determines DNA Test Accuracy

How reliable a DNA test result is depends not only on the technology used but heavily on the testing laboratory's quality control system. No matter how state-of-the-art the analysis equipment, accurate results cannot be obtained unless sample handling and data management are done properly.

1. Presence of international accreditation
   Obtaining AABB accreditation, ISO/IEC 17025, or ISO 9001 certification proves that a testing laboratory meets international quality standards. [ref:3]
2. Number of STR markers used
   The more markers used, the higher the accuracy and reliability. Current standard testing uses 20 or more markers, and some laboratories using the latest technology can analyze 24 or more.
3. Sample management system (chain of custody)
   For legal testing, a system that proves the sample was properly managed consistently from collection through to reporting is essential.
4. Privacy protection efforts
   DNA information is the ultimate form of personal information. It is important to choose an institution that takes thorough measures to protect personal information, such as obtaining Privacy Mark certification, encrypting data, and enforcing strict access controls.

Key Points for Understanding Your Results

To correctly understand your DNA test results, keep the following points in mind.

• A positive result (99.9% or higher) is very strong evidence — Treated as "effectively conclusive" even under international standards, and it carries high evidentiary weight in court.
• A negative result (0%) is usually correct, but misjudgment can occur in special cases — In extremely rare cases such as chimerism, mosaicism, or sample contamination, there is a slight possibility that a negative result could be overturned.
• Testing lab quality determines the reliability of the result — Check for AABB accreditation, ISO/IEC 17025 accreditation, the number of STR markers used, and the sample management system.
• The reason it never reaches 100% is a statistical principle — In probability theory it is theoretically impossible to prove a hypothesis with 100% certainty, but 99.99% is practically equivalent to certainty. [ref:4]
• Retesting is possible if you have doubts about a result — Accuracy can be verified through retesting with a different sample or supplementary analysis using additional STR markers.

The seeDNA Genetic Medical Research Institute provides highly accurate DNA testing based on its own strict quality control and international standards. It holds ISO 9001 international quality certification and Privacy Mark certification, and maintains a thorough system covering everything from sample handling to analysis and reporting.

Frequently Asked Questions

Q1. Can a 99.9% paternity positive probability be considered the same as 100%?

A. Statistically, it is theoretically impossible to declare a hypothesis correct with 100% certainty, but a positive probability of 99.9% or higher is considered "effectively conclusive." The international standard AABB also uses 99% or higher as its criterion for legal testing, so in practice it is appropriate to conclude that "the parties are almost certainly parent and child." The probability that an unrelated person would coincidentally match this many genetic markers is less than 1 in several hundred billion.

Q2. If a negative result (0%) comes back, can it be concluded with certainty that they are not parent and child?

A. Negative results are usually extremely accurate. However, genetic exceptions such as chimerism or mosaicism, or sample contamination or mix-ups, can, in extremely rare cases, cause an incorrect negative result. If you have doubts about a result, consider retesting using a different sample (such as blood or hair roots).

Q3. What is chimerism, and how does it affect DNA testing?

A. Chimerism refers to a condition in which two or more genetically distinct sets of DNA exist within one body. It is thought to arise when, during the fetal period, one embryo of a pair of fraternal twins is absorbed into or fused with the other. This can result in different DNA types in blood versus germ cells, meaning a mismatch (false-negative) can occur in testing even between an actual parent and child. However, it is an extremely rare phenomenon worldwide.

Q4. What are the STR markers used in DNA testing?

A. STR (Short Tandem Repeat) refers to short repeating segments of base sequences found in the human genome. Because the number of repeats differs between individuals, STRs are used for personal identification and parentage testing. Current standard DNA testing analyzes 20 or more STR markers simultaneously, achieving extremely high accuracy in determinations.

Q5. How can I choose a reliable DNA testing institution?

A. Check whether the institution holds AABB accreditation or ISO/IEC 17025 accreditation, whether it uses 20 or more STR markers, whether it maintains a chain of custody, and whether it has privacy protection measures such as Privacy Mark certification. The seeDNA Genetic Medical Research Institute holds ISO 9001 and Privacy Mark certification and conducts testing under strict quality control.

Q6. Can a retest change the result after a positive result has been obtained?

A. If a positive result of 99.9% or higher has been obtained, there is almost no possibility that a retest would overturn the result. This is because it is based on extremely strong statistical evidence — 20 or more STR markers matching without contradiction — so testing with additional markers will not change the conclusion. If a retest is required for legal proceedings, an independent test at a different accredited institution is also possible.

Q7. What is the difference between mosaicism and chimerism?

A. Chimerism is a condition in which cells originating from two different fertilized eggs coexist within one body, whereas mosaicism is a condition in which, after developing from a single fertilized egg, a mutation occurs in some cells, resulting in genetically distinct cell populations. Both can cause results to differ depending on the collection site in DNA testing, but both are extremely rare phenomena.