A centimorgan is a unit of genetic distance that measures how likely two spots on a chromosome are to be separated during the reshuffling of DNA that happens when sperm and egg cells form. One centimorgan corresponds to a one-percent chance that a crossover event will land between two genetic markers in a single generation. The name honors the geneticist Thomas Hunt Morgan, whose fruit fly experiments in the early twentieth century laid the groundwork for chromosome mapping. Despite sounding like a physical measurement, centimorgans track probability rather than actual stretches of DNA, and that distinction turns out to matter a great deal for everything from ancestry testing to hunting down disease genes.
How Centimorgans Measure Genetic Distance
When your body makes eggs or sperm, paired chromosomes line up and swap segments of DNA in a process called recombination. Each swap creates a crossover, a point where the maternal version of a chromosome switches to the paternal version or vice versa. Centimorgans quantify the frequency of these crossovers between any two positions on a chromosome. If two genetic markers sit 1 cM apart, there is roughly a 1 in 100 chance that a crossover lands between them during any single round of cell division. At 50 cM apart, the two markers behave almost as though they are on separate chromosomes, because a crossover between them is nearly guaranteed in every generation.
Recombination is not merely a random accident. It is a programmed event essential for accurate chromosome sorting during cell division, and it creates new combinations of gene variants that get passed to offspring. Without crossovers, chromosomes would not pair up and separate correctly, and the resulting egg or sperm cells would end up with the wrong number of chromosomes.
The total genetic length of the human genome is roughly 3,500 to 4,000 cM on a sex-averaged map. A landmark collaborative map placed the figure at about 4,000 cM, built from nearly 6,000 genetic markers.1PubMed. A comprehensive human linkage map with centimorgan density That means, on average, the typical person experiences around 35 to 40 crossovers spread across all their chromosomes in each generation. But “on average” does a lot of heavy lifting here, because the rate of crossover varies wildly depending on where you look in the genome and who you are looking at.
Why Genetic Distance and Physical Distance Don’t Match
One of the most counterintuitive things about centimorgans is that they do not map neatly onto physical distance measured in base pairs. You might expect that a million base pairs (a megabase) would always equal the same number of centimorgans. In reality, the ratio fluctuates dramatically. Across the human genome, recombination rates range from essentially zero to at least 9 cM per megabase, depending on the chromosomal region.2PubMed. Comparison of human genetic and sequence-based physical maps Two genes sitting a megabase apart in a recombination “hotspot” can be separated by several centimorgans, while two genes the same physical distance apart in a “coldspot” may be separated by a fraction of a centimorgan.
This unevenness was first observed in detail in simpler organisms and later confirmed across the human genome. Work on yeast chromosomes showed that crossovers were far from uniformly distributed: some intervals had significantly more crossovers than expected, others significantly fewer. The conclusion was that the underlying DNA sequence or chromosomal structure has a dramatic effect on where recombination happens.3PubMed Central. Expansions and contractions of the genetic map relative to the physical map of yeast chromosome III Humans show the same pattern at a larger scale.
Recombination tends to be suppressed near the centromere (the pinched middle of a chromosome) and elevated near the telomeres (the tips), though this rule does not hold for every chromosome.4PubMed. Variation in recombination rate across the genome: evidence and implications On top of that broad pattern, the genome contains tens of thousands of narrow hotspots where crossovers concentrate, separated by stretches of relative quiet. Large-scale analysis using population data has identified close to 90,000 putative hotspots and about 80,000 coldspots across human populations, around two-thirds of which were previously uncharacterized.5bioRxiv. Systematic Analyses of Autosomal Recombination Rates from the 1000 Genomes Project Uncovers the Global Recombination Landscape in Humans
The practical upshot is that you cannot simply convert centimorgans to base pairs with a fixed ratio. Geneticists maintain both physical maps (in base pairs) and genetic maps (in centimorgans) because they answer different questions. A physical map tells you how many nucleotides separate two markers. A genetic map tells you how often those markers get separated by inheritance. When researchers want to know whether two genes travel together from parent to child, the genetic map is what matters.
Women and Men Have Different Genetic Maps
Recombination rates differ between biological sexes in a way that has real consequences for genetic mapping. Across most of the genome, female meiosis (egg formation) produces more crossovers than male meiosis (sperm formation). The female autosomal genetic map is, as a result, substantially longer than the male map when measured in centimorgans. Male recombination, however, is elevated near the telomeres, so men have relatively higher rates at the very tips of chromosomes.6bioRxiv. Sex differences in the genetic basis of human recombination within 190,000 parent-child pairs
This sex difference means that the same physical stretch of DNA corresponds to different centimorgan values depending on whether it was passed down through the mother or the father. In genetic genealogy, this partly explains why siblings who share the same parents can show different total shared cM values depending on how many crossovers each parent’s egg or sperm cells happened to produce. It also means that published “sex-averaged” genetic maps are a compromise: they split the difference between the female and male maps. If you see a cM distance in a database without further specification, it is usually sex-averaged.
Maternal Age Changes the Number of Crossovers
The number of crossovers per egg cell is not fixed over a woman’s reproductive life. Research using large pedigree datasets has found that children born to younger mothers tend to inherit more crossovers on their maternal chromosomes than children born to older mothers. One study estimated a decrease of about half a crossover per year of maternal age, and this effect was strongest for mothers under 32.7PLoS Genetics. Age-Dependent Recombination Rates in Human Pedigrees
Other research has complicated this picture by finding a sharp increase in crossovers among mothers over about 39, possibly reflecting a breakdown in the normal regulation of recombination in aging egg cells.8Nature Communications. Escape from crossover interference increases with maternal age These findings point in somewhat different directions, and the full relationship between maternal age and recombination is still being worked out. But the takeaway is that “how many centimorgans apart” two markers are in a given family can be influenced by the mother’s age at the time she conceived, because that affects the total number of crossovers laid down during egg formation. No comparable age effect has been reliably detected in fathers.
Centimorgans in Genetic Genealogy
If you have taken a consumer DNA test from companies like AncestryDNA or 23andMe, centimorgans are the currency used to estimate how closely you and another person are related. When the testing company compares your DNA to a match’s, it looks for stretches of your genomes that are identical. These shared segments, measured in centimorgans, become the basis for predicting your likely relationship.
The logic works like this: the more recently you and a match share a common ancestor, the longer the shared DNA segments will be, because fewer generations of crossovers have had a chance to break them up. A parent and child share about 3,400 cM (essentially half the genome). Full siblings share roughly 2,600 cM on average. First cousins share somewhere around 850 cM. By the time you reach third cousins, the average drops to roughly 75 cM, and at fourth and fifth cousin distances, many pairs share little or no detectable DNA at all.
These numbers are averages, though, and the actual amount of shared DNA varies widely around them. That variation is what makes distant relationship prediction tricky. A forensic benchmarking study found that using a minimum shared-cM threshold of 7 cM correctly classified the majority of known relationships but also produced spurious matches among unrelated individuals. Raising the threshold to 12 cM reduced those false positives without losing sensitivity for closer relationships.9Forensic Science International. Benchmarking KinSNP®: A study on genetic relationship prediction for forensic applications Predictive confidence drops noticeably once you get past about fifth-degree relatives, because the expected shared DNA gets so small that noise from random population-level sharing overwhelms the signal from actual recent ancestry.
Identity-by-Descent Versus Identity-by-State
Not all shared DNA between two people comes from a recent common ancestor. Two individuals can match at a stretch of DNA simply because the variant is common in the population. Geneticists draw a distinction between identity-by-descent, where matching segments trace back to a shared ancestor in the recent past, and identity-by-state, where alleles happen to match by coincidence.10PLOS Genetics. Inference of Relationships in Population Data Using Identity-by-Descent and Identity-by-State
Consumer DNA companies and forensic tools try to filter out identity-by-state and keep only the segments that are genuinely inherited from a shared ancestor. They do this partly by setting a minimum segment length in centimorgans. Very short matching segments (say, below 5 to 7 cM) are more likely to be coincidental population-level matches rather than evidence of a recent common ancestor. Longer segments are more convincing because a genuine block of shared DNA has to survive many generations of crossovers to remain intact. The investigative genetic genealogy literature discusses this tradeoff explicitly: set the segment threshold too low and you pick up false matches, set it too high and you miss real relatives.11Forensic Science International: Genetics. Investigative genetic genealogy: Current and future directions
Ancient DNA research pushes this challenge further. When working with degraded archaeological samples, detecting shared segments becomes harder because there are more gaps in the data. Specialized tools for ancient DNA have shown that segments longer than 10 cM can be detected with high precision (above 90%) even at low sequencing coverage, while shorter segments in the 8 to 10 cM range are more error-prone.12Nature Genetics. Accurate detection of identity-by-descent segments in human ancient DNA This has opened up the ability to reconstruct family trees from burial sites thousands of years old, using centimorgans as the yardstick for how closely ancient individuals were related.
Centimorgans in Disease Gene Mapping
Before the era of whole-genome sequencing, centimorgans were the primary rulers used to find disease genes. The strategy, called linkage analysis, works by tracking whether a genetic marker tends to be inherited alongside a disease within large families. If a marker consistently travels with the disease across generations, the gene responsible for the disease must be nearby on the chromosome, close enough that crossovers rarely separate them.
As an example, the gene responsible for North Carolina macular dystrophy, a hereditary eye condition, was localized to a 1.1 cM interval between two markers on chromosome 6.13PubMed Central. North Carolina macular dystrophy: clinical features, genealogy, and genetic linkage analysis A 1.1 cM interval sounds tiny, but because genetic and physical distance do not match one-to-one, that interval still contained enough physical DNA to require further work to pinpoint the exact gene. Narrowing a disease gene to a centimorgan-scale region was nevertheless a huge advance: it told researchers which small patch of the genome to focus on.
Modern genomics has largely replaced classical linkage mapping with direct sequencing approaches, but centimorgans remain relevant. Genetic maps calibrated in centimorgans help researchers understand the patterns of variation they see in population data, and they are built into the statistical models that underpin genome-wide association studies. Even when the final output is reported in base pairs, the recombination landscape described in centimorgans shapes which associations can be detected and how finely a causal variant can be pinpointed.
Comparing Genomes Across Species
Centimorgans also show up in comparative genomics, where researchers line up the genomes of different species to see which blocks of genes have stayed together over evolutionary time. When two species diverge, crossovers and chromosomal rearrangements gradually shuffle their genomes. By measuring the lengths of chromosomal segments that remain in the same order between species, scientists can estimate how rapidly genomes have been reshuffled.
An early study comparing the human and mouse genomes estimated that the autosomal segments conserved since the two lineages diverged had an average length of about 8 cM, with individual conserved segments ranging from 1 to 24 cM.14PNAS. Lengths of chromosomal segments conserved since divergence of man and mouse These blocks of shared gene order, measured in centimorgan units, have been used to reconstruct the ancestral mammalian genome and to identify regions where chromosomal breaks tend to cluster. Because centimorgans reflect recombination probability rather than physical size, they give a biologically meaningful picture of how often a given chromosomal block gets disrupted across generations, which is exactly what matters when you are asking how well gene order has been preserved over tens of millions of years.
Mapping Functions and the Math Under the Hood
When centimorgan distances get large, the relationship between observed recombination frequency and true genetic distance becomes nonlinear. The reason is double crossovers: if two crossovers happen in the same stretch of DNA, they cancel each other out, and the markers on either end appear not to have recombined at all. This means that raw recombination frequencies undercount the true genetic distance for markers that are far apart.
Geneticists use mathematical formulas called mapping functions to correct for this. The two most common are the Haldane function, which assumes crossovers occur independently of one another along the chromosome, and the Kosambi function, which accounts for crossover interference, the tendency of one crossover to suppress additional crossovers nearby. In practice, neither function is universally “right” because the degree of interference varies across different chromosomal regions and different organisms.15Aquaculture. Centromere-linkage in the turbot (Scophthalmus maximus) through half-tetrad analysis in diploid meiogynogenetics For the small centimorgan distances that matter most in linkage mapping (below about 10 cM), the raw recombination fraction and the corrected genetic distance are nearly identical, so the choice of mapping function matters little. It becomes important mainly for longer distances and for building complete chromosome maps.
Common Misconceptions About Centimorgans
A few misunderstandings come up repeatedly, especially among people interpreting consumer DNA test results. The first is treating centimorgans as a fixed number of genes or base pairs. Because of the hotspot landscape discussed earlier, a 10 cM segment on one chromosome can contain a very different amount of physical DNA than a 10 cM segment on another. Centimorgans are probabilities, not tape measures.
A second misconception is expecting shared cM values to be the same for all pairs of the same relationship type. Two pairs of first cousins will not necessarily share the same amount of DNA. The randomness of crossover placement means there is a wide distribution of possible shared cM values for any given relationship. This is especially true for more distant relationships, where the expected value is low and the variance is high relative to that expectation. A pair of third cousins might share 120 cM or they might share 20 cM, and both are within the normal range.
A third misconception, more common in academic settings, is assuming that the human genetic map is static. As the data on maternal age and sex differences show, the recombination landscape shifts with biological context. Population-level differences in recombination rates have also been documented, meaning the “centimorgan distance” between two markers can differ slightly depending on the ancestry of the individuals being studied. Published genetic maps are based on averaged data across many families, and any individual family’s transmission pattern will deviate from those averages.
Forensic Genetic Genealogy and Centimorgan Thresholds
Centimorgans have taken on new public significance through forensic investigative genetic genealogy, the technique used to identify suspects or victims by uploading crime-scene DNA profiles to genealogy databases and looking for partial matches. The shared cM value determines what level of relationship an investigator can infer, and therefore how much traditional genealogy work will be needed to close the gap between a database match and the unidentified person.
A close match of several hundred cM might point to a second cousin, which gives investigators a manageable family tree to work through. A distant match of 20 or 30 cM might indicate a fourth or fifth cousin, requiring extensive genealogical research to find the connection. The segment-based approach compares not just the total shared centimorgans but also the distribution of individual segments, since a few long segments suggest a closer relationship than many short ones totaling the same cM value.16Forensic Science International: Genetics. Forensic genealogy—A comparison of methods to infer distant relationships based on dense SNP data Forensic labs must carefully choose their minimum cM thresholds, balancing the risk of missing real relatives against the risk of chasing false leads from coincidental matches in the population.
The success of forensic genetic genealogy in high-profile cold cases has brought centimorgans into courtrooms and news coverage, giving a once-obscure unit of genetic measurement a surprisingly prominent role in criminal justice. For the general public encountering the term for the first time through a DNA match list or a news story about a solved case, the core idea remains the same: centimorgans tell you how much DNA two people share, which in turn tells you how recently their family trees intersected.