How Many Base Pairs Are in the Human Genome?

A single copy of the human genome contains roughly 3.05 billion base pairs, according to the first truly complete sequencing effort published in 2022. But that number describes just one set of chromosomes. Because humans carry two copies of most chromosomes, the total DNA in a typical cell runs to about 6.3 billion base pairs. Which number you encounter depends on whether someone is talking about the reference genome or the full contents of your cells, and the distinction matters more than it might seem.

The 3.055 Billion Figure and How We Got It

For two decades after the Human Genome Project declared its draft complete in 2001, roughly 8% of the genome remained unreadable. Highly repetitive stretches near centromeres and telomeres defeated the sequencing technology of the time, leaving hundreds of gaps in what scientists used as their reference. In 2022, the Telomere-to-Telomere (T2T) Consortium published a gapless sequence of a human genome, designated T2T-CHM13, spanning 3.055 billion base pairs. That assembly covered every chromosome except Y, corrected errors in earlier reference versions, and added nearly 200 million base pairs of previously unresolved sequence, including 1,956 gene predictions, 99 of which appeared to code for proteins.¹

The prior standard reference, GRCh38, had served researchers well but left large parts of the genome unmapped. The T2T-CHM13 assembly filled those gaps and reconstructed regions that had been only partially resolved, giving the scientific community its first end-to-end view of the human genome.² Long-read sequencing platforms made this possible. Earlier technologies produced short fragments that struggled with repetitive DNA, but newer platforms generate reads tens to thousands of kilobases long, enough to span and resolve even the most stubborn regions of the genome.³

One Copy or Two

The 3.05 billion number is the haploid count, meaning one complete set of 23 chromosomes. Your cells, though, are diploid: they carry 23 pairs, one inherited from each parent. When you account for both copies, the total base-pair content of a single human cell is considerably larger. A detailed accounting puts the diploid male genome (46 chromosomes, XY) at about 6.27 billion base pairs and the diploid female genome (46 chromosomes, XX) at about 6.37 billion base pairs, with an overall average near 6.32 billion.⁴

The difference between males and females comes down to the sex chromosomes. The X chromosome is much larger than the Y, so a cell with two X chromosomes carries more DNA than one with an X and a Y. It is a small difference in the grand scheme, about 100 million base pairs, but it is real and measurable.

When news articles or textbooks say “the human genome is 3 billion base pairs,” they are usually rounding the haploid number. When geneticists talk about total DNA content per cell, they mean the diploid figure. Both are correct depending on context, but mixing them up can make estimates of genome size seem inconsistent when they are not.

How Much of the Genome Codes for Proteins

Only a tiny fraction of those billions of base pairs directly instructs cells to build proteins. The human genome contains about 180,000 exons, the segments of genes that get translated into protein. Collectively those exons, called the exome, add up to roughly 30 million nucleotides, or about 1% of the genome.⁵

That leaves 99% of the genome doing something other than coding for proteins. Some of that non-coding DNA has well-understood jobs: it regulates when and where genes are switched on, provides structural support for chromosomes, or encodes functional RNA molecules that never become proteins. A substantial portion consists of repetitive sequences, many of them remnants of ancient mobile genetic elements that copied themselves throughout the genome over millions of years. And some of it remains genuinely mysterious, with functions that researchers are still working to characterize. The old label “junk DNA” has largely fallen out of favor as more regulatory and structural roles have been identified, but the question of how much non-coding DNA is truly functional versus simply tolerated is still actively debated.

The Mitochondrial Genome

The base-pair counts above describe nuclear DNA, the chromosomes packed inside the cell’s nucleus. But your cells contain a second genome. Mitochondria, the organelles that generate most of a cell’s energy, carry their own small, circular DNA molecule. The human mitochondrial genome is 16,569 base pairs long and encodes 37 genes: 13 for proteins involved in energy production, 22 for transfer RNAs, and 2 for ribosomal RNAs.⁶

Compared to nuclear DNA, the mitochondrial genome is vanishingly small, less than one-millionth the size of a single haploid set. Yet it plays a significant role in cell metabolism and energy homeostasis, directly contributing to the molecular machinery of oxidative phosphorylation.⁷ Mitochondria are not self-sufficient, though; they depend heavily on proteins encoded by nuclear genes and imported into the organelle. A single cell can contain hundreds or thousands of copies of mitochondrial DNA, so the total mitochondrial contribution to a cell’s base-pair inventory, while individually tiny per molecule, is not negligible in aggregate.

Mitochondrial DNA is also inherited differently from nuclear DNA. You get it almost exclusively from your mother, since the egg cell contributes essentially all of the mitochondria to a fertilized embryo. This maternal inheritance pattern makes mitochondrial DNA a popular tool for tracing ancestry through the maternal line.

Why Your Genome Is Not Exactly the Same Size as Anyone Else’s

The reference genome is a single representative sequence, but no two people carry exactly the same number of base pairs. Structural variation, which includes deletions, duplications, insertions, inversions, and other rearrangements, encompasses millions of bases in every human genome.⁸ Some people carry extra copies of particular genes or large stretches of duplicated sequence, while others have deletions that remove tens of thousands of base pairs. These differences mean that the total base-pair count varies from person to person by millions of bases.

Even within a single person, not every cell has the same genome. Somatic mosaicism refers to the presence of two or more genetically distinct cell populations within one body, all originating from the same fertilized egg. Mutations accumulate throughout development and over a lifetime, ranging in scale from single base-pair changes to whole-chromosome gains or losses. Classification of mosaic variation spans germline versus somatic origin, size of the mutation, which tissues are affected, and functional consequence.⁹ A skin cell in your arm and a neuron in your brain, while overwhelmingly similar, may differ at dozens or hundreds of positions accumulated since the embryo’s first cell divisions.

Extrachromosomal DNA Inside Your Cells

Beyond chromosomal and mitochondrial DNA, human cells contain a third category that most people never hear about: extrachromosomal circular DNA. These are small rings of DNA floating in the nucleus, distinct from the chromosomes. They originate from chromosomal DNA and mobile genetic elements, and they carry both coding and non-coding sequences. Under some conditions, extrachromosomal circular DNA can account for up to 20% of a cell’s total DNA content.¹⁰

That 20% figure is striking, and it needs context. Extrachromosomal circular DNA levels vary enormously depending on cell type and health status. Cancer cells, in particular, tend to harbor large amounts of extrachromosomal DNA, and this has become a hot research area because these DNA circles can carry amplified copies of cancer-promoting genes that help tumors grow and resist treatment. In a healthy, non-cancerous cell, the quantity is typically much lower. Still, the existence of this DNA pool means that the total base pairs inside a cell are not limited to what you’d calculate from the chromosomes and mitochondria alone.

Genome Size Has Surprisingly Little to Do with Complexity

If you assumed that more complex organisms would have larger genomes, the human genome’s 3 billion base pairs might seem to make intuitive sense: bigger than a bacterium’s, smaller than some plants’. But that assumption falls apart quickly. Some single-celled amoebas have genomes hundreds of times larger than ours. The marbled lungfish holds the current record for vertebrates, with a genome roughly 14 times the size of the human one. Meanwhile, the pufferfish, a vertebrate with bones, a brain, and sophisticated behavior, gets by with a genome about an eighth our size.

This disconnect between genome size and biological complexity is known as the C-value paradox: the observation that the amount of DNA in an organism’s cells does not correlate with how complex that organism appears to be.¹¹ Much of the variation comes from differences in repetitive DNA, transposable elements, and other sequences that do not code for proteins. Evolution does not optimize for a lean, minimal genome the way an engineer might. It tolerates (and sometimes benefits from) vast amounts of DNA that has no immediately obvious function.

Understanding this matters because it reframes what “genome size” actually tells you. The human genome is not 3 billion base pairs because it takes 3 billion base pairs to build a human. It is 3 billion base pairs because that is the outcome of billions of years of mutation, duplication, deletion, and selection acting on a system where most of the DNA is not under strong pressure to be any particular size.

How Different Is Your Genome from a Chimpanzee’s

A comparison of roughly 1.9 million base pairs of chimpanzee DNA against corresponding human sequences found an average sequence difference of about 1.24%.¹² That number, often rounded to “about 1%,” is one of the most widely quoted figures in genetics. It means that if you lined up the two genomes letter by letter, roughly 98.8% of the positions would match.

But that 1.24% only captures single-letter substitutions. It does not account for structural differences: places where one species has an insertion, a deletion, or a duplicated stretch that the other lacks. When those are factored in, the overall divergence is larger, perhaps 4 to 5% of the genome depending on how you measure it. The rate of change also is not uniform across the genome. Some regions have diverged far more than others, and different types of mutations occur at markedly different rates. The seemingly tidy “1% different” framing, while not wrong for single-nucleotide comparisons, understates the total genetic distance between the two species.

Still, 3 billion base pairs with only low single-digit percent differences underscores just how much genetic architecture humans share with their closest living relatives, and how much evolutionary work can be done by rearranging, duplicating, and tweaking a relatively small fraction of the total sequence.

What Whole-Genome Sequencing Actually Reads

When someone has their “whole genome sequenced,” the lab typically maps their DNA against a reference genome and identifies where their sequence differs. But what the sequencer actually reads, and how completely, depends on the technology. Short-read platforms, the workhorse of most clinical sequencing, produce fragments a few hundred base pairs long. These are fast, accurate at the single-base level, and cheap enough to be widely available. Their weakness is repetitive DNA: if a region of the genome consists of the same sequence repeated many times, short reads cannot tell where in that repetitive block they belong, so those regions go unreported or are poorly assembled.

Long-read platforms, by contrast, generate reads that span tens of thousands of base pairs or more, easily traversing repetitive stretches that short reads cannot resolve.³ This capability is what made the complete T2T assembly possible. For clinical use, long-read sequencing is increasingly being adopted to detect structural variants, large deletions, and repeat expansions that short-read methods miss. If you are thinking about getting your genome sequenced and want the most complete picture, the technology behind the test matters as much as the test itself.

A “whole exome” test, by contrast, sequences only the roughly 1% of the genome that codes for proteins. It is cheaper and targeted, but it will miss variants in regulatory regions, structural changes in non-coding DNA, and repeat expansions. Both approaches read real base pairs from your cells, but they cover very different fractions of the 3 billion.

Counting the Genome in Physical Terms

Base pairs are an abstract unit, so it can help to translate the genome into physical quantities. A single haploid copy of your DNA, stretched end to end, would measure roughly two meters. Both copies together in one cell come to about four meters of DNA packed into a nucleus about six micrometers across, roughly the width of a red blood cell. The packaging system that makes this possible, coiling DNA around histone proteins into progressively tighter structures, is itself an active layer of gene regulation. Which genes get expressed depends in part on how tightly their stretch of DNA is wound up.

In terms of weight, the diploid genome in a single cell tips the scales at about 6.5 picograms, a quantity too small to see or feel but measurable with laboratory instruments.⁴ Your entire body contains roughly 37 trillion cells, most of them carrying a full diploid genome. Multiplied out, the total DNA in a human body would weigh around 50 grams, roughly the mass of a golf ball. That golf ball, if you could unwind and line up every strand end to end, would stretch from the Sun well past the orbit of Pluto.

None of these physical facts change the base-pair count, but they ground the numbers in something tangible. Three billion is hard to picture in the abstract. Two meters of molecule crammed into a space smaller than the period at the end of this sentence is a different kind of impressive.