A Short History of DNA Testing, and How Long a Test Takes Today

The photograph that changed everything looked, by Alec Jeffreys’ own account, like a smudged supermarket barcode. It was a Monday morning in September 1984, in a cramped Leicester laboratory, and the film he pulled from the developing tank showed a pattern of dark bands from a lab technician and her parents. Half the bands matched her mother. The other half matched her father. Nobody had ever seen a person’s inheritance laid out that plainly.

Four decades later, a cheek swab can go in the mail on a Tuesday and return a pie chart of continental ancestry, a list of inherited traits, and sometimes a phone call from a half-sibling nobody knew existed. The technology has become so ordinary that its strangeness is easy to forget.

This is a short history of how we got here, what the early tests could and could not do, and the honest answer to the question people ask most once they have decided to test: how long will I be waiting?

When did DNA testing start? The 1984 moment in a Leicester lab

People had been reading human inheritance long before anyone touched DNA itself. Blood typing, first described in 1901, could prove a man was not a child’s father but could rarely prove that he was. In the 1970s, tissue-typing of the immune system’s HLA markers narrowed things further. Both methods worked with proteins on the surface of cells, not with the genetic code that produced them.

The code became readable in 1977, when Frederick Sanger’s chain-termination method allowed scientists to spell out short stretches of DNA letter by letter. The National Human Genome Research Institute describes this as the founding technique of sequencing, and versions of it are still used in clinical laboratories today. Yet Sanger sequencing in the 1970s and early 1980s was slow, expensive and confined to research.

The practical turning point came on September 10, 1984. Jeffreys had been studying regions of DNA where short sequences repeat over and over, with the number of repeats differing wildly from one person to the next. When he cut DNA with enzymes, separated the fragments by size and exposed them to a radioactive probe, the result was a banding pattern unique to each individual and shared, in halves, with each parent.

He called it DNA fingerprinting, and the first real-world use was not a murder but an immigration dispute in 1985: a boy from Ghana whose right to rejoin his mother in Britain hinged on proving she was his mother. The bands proved it. So if the question is when DNA testing started, the honest answer is 1984 for the method, 1985 for the first case it decided.

Who first used DNA to solve a crime?

The first criminal case decided by DNA also came from Leicestershire, and its most important lesson is not the conviction but the acquittal that preceded it.

Two 15-year-old girls, Lynda Mann and Dawn Ashworth, were murdered in the villages of Narborough and Enderby in 1983 and 1986. Police arrested a 17-year-old local man who confessed to the second killing. Detectives asked Jeffreys to confirm that the same person had committed both. His profiles showed two things: the same man was indeed responsible for both crimes, and that man was not the suspect in custody. Richard Buckland became the first person exonerated by DNA evidence, released in November 1986.

What followed was the first DNA mass screening. Roughly 5,000 local men were asked to give blood. None matched. The case cracked only when a bakery worker was overheard boasting that he had given a sample under a colleague’s name. That colleague, Colin Pitchfork, was arrested, matched, and in January 1988 became the first person convicted of murder on DNA evidence.

Across the Atlantic, a Florida jury convicted Tommie Lee Andrews of rape in late 1987 in what is generally recorded as the first US conviction using DNA, a few weeks before the Pitchfork verdict. Within five years the technique had been argued over in courtrooms on both continents, and the arguments were about statistics as much as science, a story that belongs to the next section.

The Narborough case set a pattern that still holds: DNA is as powerful for clearing the innocent as for identifying the guilty, and it needs a comparison sample to mean anything at all.

How accurate were DNA tests in the 80s?

The 1980s method, known as RFLP (restriction fragment length polymorphism) analysis, was genuinely discriminating when it worked. Its weaknesses lay in what it demanded and how it was read.

It demanded a lot of DNA. A usable profile needed a blood stain the size of a coin or a comparable amount of semen or tissue, and the DNA had to be largely intact. Degraded, old or tiny samples often produced no result. The process itself took days to weeks, involved radioactive probes, and was performed in only a handful of laboratories worldwide.

Reading the result was the harder problem. Fragments were separated on a gel, and matching two samples meant judging by eye whether bands had migrated the same distance. Bands could shift slightly depending on how much DNA was loaded, a phenomenon known as band shifting, and laboratories had to set tolerance windows for what counted as a match. Contamination controls and blind proficiency testing were not yet standard.

Then came the statistics. A match is only meaningful if you can say how rare the pattern is in the population, and in the late 1980s the reference databases were small and the population genetics contested. American courtrooms saw this collide head-on in 1989 in a New York case where defense and prosecution scientists jointly issued a statement that the laboratory work was unreliable. The result was a wave of standard-setting: the FBI’s technical working groups, a 1992 US National Research Council report, and accreditation schemes that laboratories now take for granted.

So were 1980s tests accurate? The biology was sound. The practice was uneven, and the profession spent the next decade learning to prove its results rather than assert them.

What changed in the 1990s: PCR, short tandem repeats and national databases

Two inventions turned DNA testing from a specialist tool into a routine one. The first was the polymerase chain reaction, or PCR, conceived by Kary Mullis in 1983 and published in 1985. PCR copies a chosen stretch of DNA millions of times in a few hours, which meant a laboratory no longer needed a coin-sized stain. A few cells on a cigarette butt or a doorknob could yield a profile.

The second was a shift in what laboratories looked at. Instead of the long, variable regions Jeffreys had used, forensic scientists moved to short tandem repeats, or STRs: tiny stretches where a two-to-five-letter motif repeats a handful of times. Each STR location has a limited number of common versions, so a single one tells you little. Combine 13, then 16, then 20 or more locations, and the pattern becomes vanishingly rare in any population. Because STRs are short, they survive in degraded samples that would defeat the older method, and because PCR amplifies them, results arrive in hours rather than weeks.

Standardization followed quickly. The United Kingdom launched the world’s first national criminal DNA database in 1995. In the United States, the FBI’s Combined DNA Index System went national in 1998 with an agreed core set of STR markers, allowing a profile from a crime scene in one state to be searched against offenders in another.

The 1990s also brought DNA into the clinic in a lasting way. Genes linked to inherited breast and ovarian cancer risk were identified in 1994 and 1995, and testing for them became one of the first widely offered predictive genetic tests. The template of a laboratory result delivered alongside genetic counseling dates from this period.

Did DNA tests exist in 2000?

They did, and in more forms than many people assume. Forensic STR profiling was routine in most developed countries. Paternity testing had moved from HLA tissue typing to DNA and was available through accredited laboratories, often ordered by courts and increasingly bought privately. Clinical laboratories offered targeted tests for specific inherited conditions, including cystic fibrosis carrier screening and the hereditary cancer genes identified in the mid-1990s. Newborn blood-spot screening, which had existed since the 1960s for a single metabolic condition, was expanding to include DNA-based confirmation for some disorders.

What did not yet exist was the whole picture. In June 2000, scientists announced a working draft of the human genome, the product of an international effort launched in 1990. The NIH’s National Human Genome Research Institute records the finished sequence as arriving in April 2003, roughly 13 years after the project began, at a cost of about 2.7 billion dollars in 1991 money. Until that reference existed, testing meant looking at genes you already knew to look at.

Ancestry testing had a faint pulse in 2000. A few companies offered analysis of the Y chromosome (passed father to son) or mitochondrial DNA (passed mother to child), which could trace a single paternal or maternal line deep into the past but said nothing about the other thousands of ancestors in a family tree. The autosomal tests that estimate continental percentages and find living relatives across the whole genome only became commercially viable after 2007, when genotyping chips could read hundreds of thousands of variants cheaply at once.

So a person in 2000 could learn whether they carried a specific disease variant or whether a man was their father. They could not learn what they would learn today from a mail-order kit.

How the cost of reading a genome fell from $100 million to under $1,000

Few numbers in medicine have fallen as fast. According to the National Human Genome Research Institute’s cost tracking, sequencing a human genome cost on the order of 100 million dollars in 2001. By 2007 it was still several million. Then a new generation of machines that read millions of fragments in parallel arrived, and the curve broke. NHGRI’s data show the cost dropping below 1,000 dollars in the second half of the 2010s, a decline that outpaced the famous Moore’s law for computer chips by a wide margin.

The mechanism deserves a plain explanation. Sanger sequencing reads one fragment at a time in one tube. Next-generation sequencing shatters DNA into millions of pieces, attaches each to a surface, and photographs all of them being copied simultaneously, letter by letter. Software then stitches the reads back together against the reference genome that the Human Genome Project produced. The bottleneck moved from chemistry to computing.

Cost, though, is not the same as price. What a clinical laboratory charges includes sample handling, quality control, interpretation by trained scientists, and the counseling that should accompany any medically meaningful result. What a consumer kit charges reflects a different product entirely: most read a few hundred thousand pre-selected variants on a genotyping chip rather than sequencing anything, which is why they can be sold for the price of a nice dinner.

The falling cost matters for one further reason. It has made it possible to sequence not just one suspected gene but a whole exome or genome in a single test, which changes both what a test can find and how long it takes to make sense of it.

What kinds of DNA tests exist today?

The phrase DNA test covers at least eight quite different things, and knowing which one you are talking about answers most questions about accuracy, cost and turnaround. MedlinePlus and the Mayo Clinic group them broadly as follows:

• Diagnostic testing confirms or rules out a suspected genetic condition in someone who already has symptoms.
• Predictive and presymptomatic testing looks for variants that raise the risk of a condition that has not yet appeared, such as certain inherited cancer or heart rhythm syndromes.
• Carrier testing tells prospective parents whether they carry a variant that could be passed to a child, often relevant when both partners carry the same one.
• Prenatal testing examines fetal DNA, either from the mother’s blood or from a sample of placental or amniotic fluid.
• Newborn screening tests a blood spot from a baby’s heel in the first days of life for conditions where early treatment changes outcomes.
• Pharmacogenomic testing looks at variants that affect how a person processes certain medicines; how any result is used is a decision for the prescribing clinician.
• Forensic and identity testing, including paternity, uses STR markers to match or exclude individuals.
• Direct-to-consumer testing is bought without a clinician and mostly reports ancestry, traits and a limited set of health-related variants.

The samples differ too. Clinical tests usually use blood, though cheek swabs and saliva are common. Prenatal tests may need a procedure. Forensic work uses whatever a scene provides. And the laboratories differ most of all: clinical tests in the United States must be performed in laboratories certified under federal standards, whereas MedlinePlus notes that consumer tests vary in what they analyze and how they report it.

How long does a DNA test take today?

The single most searched question about DNA testing has an answer that depends almost entirely on which test you mean. The laboratory chemistry is rarely the slow part. Interpretation, confirmation and the queue in front of you are.

Type of test	Typical wait for results	What drives the timeline
Clinical genetic test (single gene or panel)	A few weeks to a few months, per MedlinePlus and the NHS	Variant interpretation, confirmation, laboratory workload
Whole exome or genome sequencing	Usually toward the longer end of that range	Volume of data to analyze and classify
Consumer ancestry or trait kit	Several weeks, per MedlinePlus	Shipping, batching, matching against reference panels
Newborn blood-spot screening	Reported quickly, because some conditions need action within days	State program logistics and confirmatory testing
Forensic STR profile	Hours of bench time; weeks to months in practice	Case backlogs and evidence handling

MedlinePlus, in its guidance on how genetic testing is done, states that results for clinical tests can take from a few weeks to a few months. The NHS gives the same broad range for testing arranged through its services and adds that the referring clinician will explain the expected wait. For direct-to-consumer kits, MedlinePlus describes results arriving in several weeks, typically through an online account.

Two things shorten the wait in practice. Testing a single known variant already found in a relative is faster than searching a whole gene, because the laboratory knows exactly where to look. And in urgent settings such as a newborn intensive care unit, hospitals increasingly use rapid sequencing pathways; the evidence for how quickly these deliver actionable results is still accumulating, so treat any specific promise with caution.

Two things lengthen it. An unexpected or ambiguous variant may need a second method to confirm it. And any meaningful clinical result should be returned with counseling, which is a conversation, not a download.

Why does a fast machine still mean a slow answer?

A modern sequencer can read a whole human genome in a day or two. PCR-based tests finish in hours. So why does a patient wait weeks? Because reading is not the same as understanding.

Every genome contains millions of differences from the reference sequence. The overwhelming majority are harmless variation, the ordinary spelling differences that make us individuals. A clinical laboratory’s task is to find the handful, if any, that explain a patient’s condition or raise a specific risk. That involves filtering against population databases, checking published evidence for each candidate, and assigning a classification that ranges from benign through variant of uncertain significance to pathogenic. Two trained scientists usually review the call before it is signed out.

Uncertain results add time. When a variant has never been described, laboratories may test parents to see whether the child inherited it or it arose new, which can point toward or away from it being the cause. Confirmation by a second method is common for anything that will change care.

Then there is the queue. Genetic laboratories are specialist services, and demand has grown faster than the workforce trained to interpret results. The NHS is candid that waits vary by test and by how urgent the clinical question is.

Consumer kits look faster partly because they skip most of this. A genotyping chip reads a fixed set of variants, software compares them to reference groups, and the report is generated automatically. That is efficient. It is also why MedlinePlus advises that health-related findings from such kits should be confirmed in a clinical laboratory before anyone acts on them.

Why is DNA testing banned in France? (It isn't, quite)

The claim circulates widely, and it is half right. France does not ban DNA testing. It restricts who may order it and why.

Under the French Civil Code, genetic identification of a person may be carried out only for medical or scientific research purposes, or in the course of legal proceedings on the order of a judge. Paternity testing outside a court case is not permitted, and neither is sending a saliva sample abroad for a recreational ancestry report. Individuals who do so risk a fine, and companies are barred from marketing such kits to French residents. The country’s bioethics laws, revised several times since 1994, frame this as protecting the peace of families and preventing genetic information from being collected without a clear purpose.

France is unusual in how far it goes, but the underlying instinct is not unusual. Germany’s genetic diagnostics law requires that predictive health testing be ordered and explained by a physician, and prohibits covert paternity testing on a child without both parents’ consent. Several other European countries limit health-related consumer testing while permitting ancestry kits.

The United States takes a different route. Anyone may buy a kit, but the Genetic Information Nondiscrimination Act of 2008 prohibits health insurers and most employers from using genetic information against a person. MedlinePlus notes that this protection does not extend to life, disability or long-term care insurance, a gap worth knowing about before testing.

Whatever one thinks of the French approach, it answers a real question: who owns the information in a cheek swab, and who gets to decide what it is used for? Those questions are the subject of the next two sections.

How accurate are DNA tests today?

Accuracy has three layers, and MedlinePlus separates them usefully. Analytical validity asks whether the test correctly reads the DNA letters. Clinical validity asks whether the variant it finds is genuinely linked to the condition. Clinical utility asks whether knowing the result changes anything a person or clinician would do.

On the first layer, modern clinical sequencing performs extremely well; laboratories run known controls alongside patient samples and confirm unexpected findings. Where errors occur, they tend to arise from sample mix-ups or from regions of the genome that are technically hard to read, not from the chemistry itself.

The second layer is where honesty matters most. A variant classified as pathogenic today may be reclassified as more evidence accumulates, and the reverse also happens. Variants of uncertain significance are common in broad panels, and they are not a diagnosis. For ancestry, MedlinePlus explains that estimates depend on the reference populations a company uses; the same DNA can yield different percentages from different services, and results are revised as reference panels grow.

The third layer is the one most often skipped. Finding a variant that raises risk of a common condition by a modest amount rarely changes advice that already applies to everyone. Finding a variant in a gene with established management guidelines can change screening from a certain age onward. The difference lies in the gene, not the technology.

Forensic and paternity STR testing, when performed by accredited laboratories with proper controls, is regarded by courts as highly reliable. The historical cases of error have almost always involved handling, contamination or interpretation of mixed samples rather than the method.

Who else learns something when you spit in a tube?

DNA is the only medical information that is never entirely your own. Half of it is your mother’s, half your father’s, and roughly half is shared with each sibling and each child. A test on one person is, unavoidably, a partial test on relatives who never consented.

This has practical consequences. Consumer relative-matching has revealed misattributed parentage, undisclosed adoptions and donor conceptions to people who were not looking for them. Genetic genealogy has also identified suspects in cold criminal cases through distant cousins who uploaded profiles to public databases, a practice that has raised debate about consent even where it has solved crimes.

In clinical settings, laboratories and counselors handle this deliberately. Before predictive testing, a counselor will usually discuss whom a result might affect and how a person might share it. Some families choose to test the oldest affected relative first, so that a specific variant is known before younger relatives decide whether to look for it.

Privacy terms deserve a slower read than most people give them. MedlinePlus advises checking whether a company will store your sample, share de-identified data with researchers or third parties, and what happens to your data if the company is sold. Deleting an account does not always destroy a stored sample.

None of this argues against testing. It argues for treating a DNA result as a family document rather than a personal one, and for deciding in advance what you would do with a surprise. The people who fare best in follow-up studies tend to be those who thought about that question before opening the envelope.

Should you take a DNA test? An evidence-first opinion

Here is where a careful magazine writer is allowed a view, provided it rests on the evidence. The view is this: the value of a DNA test is decided before the sample is taken, by the question you are asking.

If the question is medical, the strongest path runs through a clinician. A family history of the same cancer in several relatives, a sudden cardiac death in a young relative, or a known variant already identified in the family are the situations where guideline bodies in the United Kingdom, the United States and Europe recommend assessment and, where appropriate, testing with counseling. The test is chosen to fit the question, the laboratory is accredited, and someone is on the other end of the phone when the result arrives.

If the question is about origins or curiosity, a consumer kit answers it reasonably well, with two caveats that MedlinePlus states plainly. Ancestry percentages are estimates that will shift as reference data grow, and any health-related finding should be confirmed clinically before it changes a decision. A negative result on a consumer kit is also not reassurance, because most kits look at only a small selection of the variants in any given gene.

What the evidence does not support is testing everyone for everything. Broad screening of healthy people without a specific indication produces many uncertain findings for every actionable one, and the anxiety and follow-up testing that result are real costs. Professional bodies have consistently declined to recommend population-wide genome sequencing for this reason.

So: ask the question first. Then choose the test that can actually answer it.

When should you talk to a clinician about genetic testing?

Genetic testing is rarely an emergency, but certain patterns in a family or a person’s health deserve a prompt conversation rather than a mail-order kit. The NHS and the Mayo Clinic both describe referral to a genetics service as appropriate when a condition appears to run in the family, when a relative has been found to carry a specific variant, or when a child has unexplained developmental or physical features.

Red-flag patterns worth raising with a doctor include:

• The same cancer, or related cancers such as breast and ovarian or bowel and uterine, in two or more close relatives, especially before age 50.
• A close relative who died suddenly and unexpectedly at a young age, or who has fainted repeatedly during exercise, which can point to an inherited heart rhythm or heart muscle condition.
• A known disease-causing variant already identified in a parent, sibling or child.
• A child with delayed development, unusual growth, seizures or several birth differences without an obvious explanation.
• Two pregnancy losses or more, or a previous child with a genetic condition, when planning another pregnancy.

Some symptoms need urgent care regardless of any genetic question. Fainting during exertion, chest pain, a new seizure, or a baby who is unusually floppy, feeding poorly or vomiting persistently in the first days of life should be assessed the same day; the genetic conversation can follow.

A genetic counselor or clinical geneticist can review a family history, explain what a test could and could not show, and estimate how long results will take in your setting. If you have already taken a consumer test and found something worrying, bring the raw report; clinicians would far rather confirm or correct a finding than have you carry it alone.

Frequently asked questions

When did DNA testing start?

Modern DNA testing began in September 1984, when Alec Jeffreys developed DNA fingerprinting at the University of Leicester. Earlier methods such as blood typing (from 1901) and HLA tissue typing (1970s) could exclude relationships but rarely confirm them. The first use of Jeffreys’ method was a 1985 immigration case; the first criminal conviction came in 1988. DNA sequencing itself dates to Frederick Sanger’s 1977 method.

How accurate were DNA tests in the 80s?

The biology was sound but the practice was uneven. 1980s RFLP tests needed a coin-sized blood stain of intact DNA, took days to weeks, and were read by eye from gel bands that could shift slightly. Population databases for calculating match rarity were small, and courtroom disputes in 1989 exposed laboratory inconsistencies. Standard-setting through the early 1990s, including accreditation and agreed statistical methods, addressed most of these weaknesses.

Who first used DNA to solve a crime?

Alec Jeffreys, working with Leicestershire police, first used DNA profiling in the investigation of two murders in Narborough and Enderby, England. In 1986 it exonerated a 17-year-old who had confessed, and in January 1988 Colin Pitchfork became the first person convicted of murder on DNA evidence after a mass screening of about 5,000 men. In the United States, a Florida rape conviction in late 1987 is usually recorded as the first.

Did DNA tests exist in 2000?

Yes. By 2000, forensic STR profiling, court-ordered and private paternity testing, and clinical tests for specific inherited conditions such as hereditary breast and ovarian cancer genes were all available through laboratories. A draft of the human genome was announced that June, with the finished sequence following in April 2003. Whole-genome ancestry kits with relative matching did not become commercially practical until after 2007.

Why is DNA testing banned in France?

It is restricted rather than banned. Under the French Civil Code, genetic identification may be carried out only for medical or scientific purposes or by order of a judge in legal proceedings. Private paternity tests and recreational ancestry kits therefore are not permitted, and individuals who use them risk a fine. France’s bioethics laws frame this as protecting families and preventing genetic data from being collected without a defined purpose.

How long does a clinical genetic test take?

MedlinePlus and the NHS both state that clinical genetic test results usually take from a few weeks to a few months. Testing for a single variant already known in a relative is generally faster than sequencing a whole gene or panel, because the laboratory knows exactly where to look. Unexpected or uncertain findings may need confirmation by a second method, which adds time. Your referring clinician can give the expected wait for your specific test.

How long does a consumer ancestry DNA kit take?

MedlinePlus describes direct-to-consumer results as typically arriving in several weeks, usually through an online account. The timeline includes mailing the sample, laboratory batching, genotyping on a chip that reads a fixed set of variants, and comparing the result against reference populations. Ancestry percentages are estimates and may be revised as a company’s reference data grow, so results can change over time.

Why does a DNA test take weeks if the machine only needs a day?

Because reading DNA is faster than interpreting it. A genome contains millions of differences from the reference sequence, almost all harmless. Clinical scientists filter these against population databases and published evidence, classify each candidate variant, and often confirm important findings by a second method before two reviewers sign off. Laboratory queues and the counseling appointment to return results add further time. Consumer kits skip most of these steps.

Are DNA test results always right?

Modern clinical sequencing reads DNA letters with very high reliability, but a correct read is not the same as a settled meaning. Variant classifications can change as evidence accumulates, and variants of uncertain significance are common on broad panels. Ancestry estimates depend on reference populations and differ between services. MedlinePlus advises confirming any health-related consumer finding in a clinical laboratory before acting on it.

Who should consider genetic testing through a doctor?

Guideline bodies suggest assessment when the same or related cancers appear in two or more close relatives, especially before age 50; when a young relative died suddenly or faints during exercise; when a specific variant has already been identified in the family; or when a child has unexplained developmental or physical differences. A genetics service can match the right test to the question and explain what it can and cannot show.

References

• How is genetic testing done? – MedlinePlus Genetics
• How is genetic ancestry testing done? – MedlinePlus Genetics
• Genetic and genomic testing – NHS

This article is for general information only and is not a substitute for professional medical advice. Please consult a qualified doctor about your individual situation.