DNA Paternity Test: How It's Done and How Accurate It Is

A DNA paternity test is a genetic test that determines whether a specific man is the biological father of a child. The method is simple: a child receives approximately half of their genetic material from their biological mother and half from their biological father. Therefore, the laboratory compares sections of the child's DNA with sections of the DNA of the alleged father and, if possible, the mother. [1]

Modern research typically relies on the analysis of multiple short, repetitive DNA sequences. The laboratory determines the genetic profile of each participant and then assesses whether the alleged father can explain the DNA variants the child could not have inherited from the mother. This is the basis for calculating the statistical strength of the match. [2]

In practice, cells from the inner cheek are most often used for this type of analysis. This is a standard, non-invasive, and convenient method of collecting the sample. Official sources also indicate that a properly collected swab from the inner cheek is comparable in accuracy to a blood test, so blood is usually not needed for routine testing. [3]

It's crucial to understand that the test answers the question of the child's biological origin. It does not determine who the social father was, who raised the child, who is the legal father according to the laws of a particular country, or what family responsibilities have already arisen under the law. In complex family situations, biological, social, and legal paternity may not coincide. [4]

From a medical perspective, such testing is important not only because of disputes over a child's parentage. Establishing a biological relationship can help clarify family history, assess the risks of hereditary diseases, and better understand the family's medical history. However, a paternity test alone does not replace a full examination for hereditary diseases and does not answer the question of whether a child has a specific genetic disorder. [5]

Table 1. What the test shows and what it does not show

Question	Can answer	Can't answer
Is the man the biological father of the child?	Yes	-
Do certain sections of the child's and the man's DNA match?	Yes	-
Who is the legal father according to law?	Partially, only together with legal procedure	Yes, the test itself doesn't solve this.
Who raised the child and performed the parental role?	No	Yes
Does the child have a specific hereditary disease?	No, other research is needed for this.	Yes
Is it possible to immediately create a complete family medical prognosis based on the test?	No	Yes

Source for the table. [6]

When is the test performed and what samples are used?

Paternity testing is most often performed in three groups of situations. The first is personal, when the family wants clarity. The second is legal, when the issue concerns child support, guardianship, inheritance, immigration procedures, or document changes. The third is medical, when the child's origin is important for family history and risk assessment for certain conditions. [7]

After a child's birth, testing is typically performed using cell samples from the buccal mucosa. This is the most common option. However, genetic testing in general can also use other samples: blood, hair, skin, amniotic fluid, and other tissues. For standard paternity testing, a buccal swab is usually preferred because it is simple, safe, and well-suited for routine laboratory work. [8]

If the alleged father is unavailable, deceased, or refuses to provide material, it is sometimes possible to completely resolve the issue by analyzing close biological relatives: the man's parents, brothers, sisters, and other blood relatives. However, this is no longer a classic direct paternity test, but a more complex kinship study. Its statistical power is usually lower, and interpretation requires greater caution. [9]

For the laboratory, the best design is considered to be a study involving three participants: the mother, the child, and the alleged father. Having a mother's sample makes it easier for the laboratory to separate the maternal portion of the child's genome from the paternal portion and more accurately estimate the contribution of the alleged father. If the study is performed on only two participants, the evidentiary strength remains high in many cases, but the risk of difficult or less confident interpretation increases. [10]

Non-standard samples, such as a toothbrush, razor, cigarette butt, chewing gum, or hair, may theoretically contain useful DNA, but they present more challenges. The main difficulties include contamination, mixing with foreign biological material, uncertainty about the identity of the object, and a lower suitability for legally significant conclusions. Therefore, for reliable, and especially forensic, testing, such samples are considered a backup rather than a primary option. [11]

Table 2. Which samples are used most often?

Sample	How convenient is it?	How typical is it?	Comment
A smear from the inside of the cheek	Very convenient	Basic	Standard for most tests
Blood	Convenient, but more invasive	Less often	Comparable in accuracy to a cheek swab when collected correctly
Hair from the root	Limited convenience	Rarely	The quality of the material can be unpredictable
Amniotic fluid	Only with prenatal invasive testing	Special case	Used in special circumstances
Placental tissue	Only with prenatal invasive testing	Special case	Not suitable for routine postpartum testing
Household items	Low predictability	Non-standard option	Increases the risk of controversial results and questions about authenticity

Source for the table. [12]

Table 3. Study for 3 participants and for 2 participants

Scheme	Who is participating?	Strengths	Restrictions
Full 3-member circuit	Mother, child, alleged father	The most powerful and easy to interpret	Samples from all 3 people are needed.
2-Participant Scheme	Child and alleged father	Possible if mother is unavailable	Higher risk of complex statistical assessment
Indirect kinship research	The child and relatives of the alleged father	Useful if the father himself is unavailable	A less direct and often less convincing path

Source for the table. [13]

How is the analysis performed and how are the results read?

After collecting the sample, the laboratory isolates DNA, identifies genetic variants in each participant, and compares them. A statistical indicator is then calculated that shows how much better the obtained data is explained by the hypothesis of biological paternity than by the hypothesis of no such relationship. In forensic genetics, this indicator is considered a likelihood ratio, and the paternity index is a special version of this ratio for the task of establishing paternity. [14]

The final report typically represents not the raw genetic data, but rather the interpreted result. In practice, laboratories most often formulate the conclusion either as ruling out paternity or as a very high probability of paternity. Clinical sources explicitly state that in everyday reports, 0% is often used for ruling out paternity and 99.9% is used for confirmatory results, although the exact format of the report depends on the laboratory and the methodology. [15]

However, a beautiful number alone doesn't mean all laboratories and agencies will interpret the result identically. For example, in US immigration practice, a result of 99.5% or higher is considered sufficient to establish a biological relationship, while the National Institute of Justice's training materials indicate that the thresholds applied may vary between countries and assessment systems. This means that it's not just the percentage that matters, but also the context, laboratory accreditation, and the quality of the sample collection procedure. [16]

How the laboratory handles mismatches is also important. In general kinship guidelines, multiple obvious mismatches usually lead to exclusion, but a single problematic spot does not always mean the man is definitely not the father. In rare cases, mutations in the DNA region being analyzed are the cause, so difficult cases require a broader panel of markers and a more careful expert assessment, rather than a mechanical conclusion based on a single questionable region. [17]

The key practical conclusion follows from this: the phrase "the test is 99.9% accurate" oversimplifies reality. Reliability depends on the quality of the samples, the number and type of markers tested, the mother's participation, the absence of a close relative among the possible fathers, and the correct statistical calculations. Therefore, a good test is not only a matter of laboratory technology but also the proper organization of the entire procedure, from sample collection to the issuance of a conclusion. [18]

Table 4. How the result is usually interpreted

Output option	What does it mean
Paternity is excluded	Genetic data are inconsistent with biological paternity
Paternity is not excluded, the probability is very high	Genetic evidence strongly supports biological paternity
The result is not convincing enough	Additional markers, maternal sample, or repeat testing are needed.
The result is controversial due to the complex case	Possible close relationship of the alleged fathers, mutation, rare biological phenomenon or sample problem

Source for the table. [19]

Home, legally valid and prenatal testing

For personal use, there's a so-called at-home format: participants collect their own samples and send them to the lab. This option is convenient for private investigations, but its main drawback is the impossibility of reliably proving who provided the samples. Therefore, such results are usually unsuitable for court or other official proceedings. [20]

Legally significant research is structured differently. Samples are collected at an approved collection point, the identity of participants is confirmed with documents, and the movement of the material is documented from collection to the issuance of a report. Professional standards for kinship laboratories and official government procedures rely on precisely this: not only the accuracy of analysis, but also the continuous, documented preservation of samples. [21]

Prenatal paternity testing is possible. The non-invasive method is based on the fact that free fetal DNA circulates in the pregnant woman's blood. Official sources on free fetal DNA indicate that such a blood test can be performed as early as the 10th week of pregnancy, and modern reviews confirm that prenatal paternity testing methods are also being developed on this basis. [22]

Invasive prenatal options also exist. Chorionic villus sampling (CVS) is typically performed between 10 and 13 weeks, and amniocentesis is performed between 15 and 20 weeks. However, both methods carry a risk of pregnancy loss: for amniocentesis, the UK National Health Service cites a risk of up to 1 in 200 procedures after 15 weeks, while for CVS, it is less than 1 in 200 for most singleton pregnancies. [23]

Therefore, in modern practice, prenatal paternity testing requires a particularly balanced approach. If the question concerns only the establishment of biological paternity, rather than medical diagnosis of the fetus for obstetric reasons, it is logical to first discuss a non-invasive option and consult a genetic specialist. Invasive methods are best considered only after a separate discussion of the risks, benefits, and alternatives. This is not a prohibition, but a cautious clinical conclusion based on the fact that safe and risky options differ fundamentally. [24]

Table 5. Personal and legally significant test

Characteristic	Personality test	Legally significant test
Who collects samples?	Usually the participants themselves	Authorized specialist
Identity verification	Limited or absent	Mandatory
Documenting the movement of samples	Usually no	Yes
Suitable for court	Usually no	Yes, if jurisdictional requirements are met
The main goal	Personal clarity	Official proof

Source for the table. [25]

Table 6. Options for prenatal paternity establishment

Method	When is it possible?	Material	Pros	Restrictions
Non-invasive blood test for pregnant women	Usually from the 10th week	Free fetal DNA in the blood of a pregnant woman	Safer for pregnancy	Methodologies vary between laboratories
Chorionic villus sampling	Approximately 10-13 weeks	Placental tissue	Early term	Invasiveness and risk of pregnancy loss
Amniocentesis	Usually 15-20 weeks	Amniotic fluid	High diagnostic value of the material	Invasiveness and risk of pregnancy loss

Source for the table. [26]

Limitations of the method, difficult cases and practical conclusions

Although paternity testing is considered highly accurate, it is not a magical test that works flawlessly in every situation. Cases where there are close relatives among the possible fathers are particularly challenging. Population-based calculations have shown that the risk of false positives is significantly higher when testing only the child and the alleged father than when testing the entire test including the mother. [27]

A separate group of rare but important problems arises from unusual biological situations. Cases of chimerism have been described, in which a single individual has two genetically distinct cell lines. In such a situation, DNA from a cheek swab or blood sample may not match the DNA from the germ cells, and a standard test can erroneously exclude the true biological father. This is extremely rare, but it is precisely these exceptions that explain why, in controversial cases, repeat testing on other tissues and a more extensive examination are sometimes requested. [28]

It's equally important to remember that biological origin is only one aspect of the issue. In adoption, the use of donor gametes, surrogacy, and certain other circumstances, social and legal parental roles may not coincide with genetic ones. For the laboratory, biological relationship is key, but for the family and the court, this is sometimes insufficient to resolve the entire dispute. [29]

Genetic testing carries not only laboratory but also psychological burdens. General information on the risks of genetic testing notes that the results can cause anxiety, guilt, depression, and tension within the family. In the case of paternity tests, the consequences often affect trust, financial obligations, and family relationships. Therefore, it makes sense to test not on impulse, but after understanding exactly what will be done with the results. [30]

The most sensible approach seems to be to first define the purpose of the study, then choose a direct or indirect route, include a maternal sample if possible, use only an accredited laboratory with documented sample flow for official purposes, and discuss the safety of the method separately during pregnancy. If the result is unexpected or conflicts with obvious circumstances, what's needed is not a scandal, but a re-evaluation by experts of the sample quality, the composition of the participants, and rare biological exceptions. [31]

Table 7. Situations that most often complicate interpretation

Situation	Why is it important?
Study without maternal sample	Reduces statistical power compared to the full design
Possible fathers include a brother or other close relative.	Increases the risk of false activation
Unexpected mismatches in 1 DNA region	May be associated with a mutation and require further testing.
Rare biological phenomena, such as chimerism	They may give a false exception.
Unconfirmed sample origin	Makes the conclusion legally vulnerable
Using non-standard objects instead of a standard brushstroke	Increases the risk of contamination and disputes over the ownership of the material

Source for the table. [32]

Frequently Asked Questions

How accurate is a paternity test?
With proper sample collection and laboratory procedures, the test is considered highly accurate. Clinical reports often cite a 99.9% result for a confirmatory conclusion, while official US immigration practice considers 99.5% or higher sufficient to establish a biological relationship. However, accuracy depends not only on the laboratory, but also on who is involved in the test and how the samples were collected. [33]

Is it possible to conduct a test without the mother?
Yes, it is possible, but it is a less powerful option than a full design with the mother. A study with two participants often leaves more room for complex statistical interpretation, especially if there are close relatives among the possible fathers. [34]

Is it possible to perform a test without identifying the alleged father?
Sometimes yes, by examining his close blood relatives. However, this is no longer a classic direct paternity test, but a more complex kinship study, where the conclusion is usually less direct and more dependent on the composition of the available relatives. [35]

Is a cheek swab sufficient, or is blood required?
In most cases, a cheek swab is sufficient. Official sources indicate that a properly performed cheek swab is comparable in accuracy to a blood test. This is why it is considered the standard material for most paternity tests. [36]

Will a court accept a home test?
Usually not. Official use requires verification of the participants' identity and documented integrity of the samples from collection to laboratory confirmation. A home kit typically does not provide this. [37]

Is it possible to establish paternity during pregnancy?
Yes. The non-invasive approach is based on free fetal DNA in the mother's blood and is usually possible from the 10th week. Invasive options also exist, but they are associated with the risk of pregnancy loss and therefore require separate discussion. [38]

Which is better during pregnancy: a non-invasive or invasive option?
If the goal is to establish paternity rather than mandatory prenatal medical diagnosis for obstetric indications, in practice it is usually wiser to first discuss the non-invasive option. Invasive procedures are associated with procedural risks and therefore require particularly careful decision-making. [39]

Can the test be wrong if the alleged father is the real father's brother?
Yes, these are the most difficult cases. Close relatives share more DNA than strangers, so the risk of false positives increases, especially if the mother's sample is not available. [40]

Is there a false exclusion of a true father?
Rarely, but it does occur under special circumstances. Among the causes described are mutations in the studied regions and very rare phenomena such as chimerism, when the genetic makeup of different tissues of one person does not completely match. [41]

Can a pedigree test be used instead of a paternity test?
No. Pedigree tests and paternity tests serve different purposes. To confirm biological paternity, a specific ancestry test is needed, not a general consumer ancestry test. [42]

Does such a test have any medical benefit for the child?
Yes, in some cases. Establishing biological origin helps to more accurately gather a family history and understand whether there are any hereditary conditions in the family that the doctor should be aware of. However, this does not make a paternity test an automatic test for diseases. [43]

Conclusion

DNA paternity testing is a highly accurate genetic test that works best when the purpose is clear in advance, samples are collected correctly, and the laboratory adheres to professional standards. The most convincing method is a direct analysis involving three participants: the mother, the child, and the alleged father. For official procedures, not only laboratory accuracy but also the documented authenticity of the samples is crucial. [44]

The method's main limitations stem not from the fact that "DNA can lie," but from the fact that real family situations can be more complex than a simple formula. Close relatives among potential fathers, the absence of a maternal sample, rare mutations, chimerism, the use of non-standard objects, and the confusion of biological, social, and legal paternity can dramatically complicate interpretation. Therefore, the best test is not the most flashy promotional package, but a properly organized and professionally interpreted examination. [45]