Traits pass from parents to offspring through DNA, the chemical instruction set packaged into chromosomes. Each parent contributes one copy of every gene, and the combination of those two copies, plus chemical switches that control how they are read, determines much of an offspring's biology. Sperm and egg each carry half the usual chromosome number. At fertilization those halves pair up, restoring a full set in the new cell. From that single cell, repeated division and gene activation build a whole organism.

This explainer goes deeper than the standard summary. It covers the molecular mechanics, the numbers behind them, the cases where the rules bend, and where inherited risk stops and environment begins.

The physical mechanism of transmission

Inheritance starts with meiosis, the specialized cell division that makes gametes. A human body cell carries 46 chromosomes: 23 pairs, one set from each parent. Before meiosis, each chromosome is copied, so the cell briefly holds 92 chromatids. The cell then divides twice, producing four gametes with 23 chromosomes each. When sperm and egg fuse, the resulting zygote returns to 46.

The pairing is not random. During the first meiotic division, matching chromosomes line up and exchange segments in a process called crossing over. Human chromosomes average roughly one to three crossovers per pair per meiosis, which reshuffles alleles so that each gamete carries a novel combination. This is why full siblings resemble each other but are never identical, except identical twins. Recombination is the reason you can inherit your grandmother's nose and your grandfather's hair from the same chromosome.

Each gene occupies a fixed locus on a chromosome. Alternative versions of a gene at that locus are alleles. You inherit one allele from each parent for most genes. Which allele gets expressed depends on dominance relationships and on regulatory signals that can silence one copy entirely.

A useful plain-language frame: genes are recipes, alleles are variant versions of a recipe, and chromosomes are the cookbooks that hold them.

Dominant, recessive, and the numbers behind them

For a simple dominant-recessive trait, the expression follows a predictable probability. Two carrier parents, each with one dominant and one recessive allele, have a 25 percent chance of a child who is homozygous recessive. That child shows the recessive trait. There is a 75 percent chance the child shows the dominant trait, though two of those three share carriers who do not show it.

These figures come from Mendel's laws, derived from pea-plant experiments in the 1860s and confirmed in organisms from fruit flies to humans. The math is clean for single-gene traits, which are rare. Most human traits that matter medically, such as height or blood pressure, are polygenic: hundreds or thousands of small-effect variants combine, so the offspring's value clusters near the parents' average but can fall anywhere in a wide range.

Cystic fibrosis illustrates a single-gene case with real numbers. It is autosomal recessive, caused by variants in the CFTR gene. Two carrier parents face a 25 percent chance of an affected child per pregnancy. Roughly 1 in 30 Americans of European ancestry carries one CFTR variant, per CDC estimates. That prevalence implies about 1 in 3,600 births affected among that population, before screening reduces observed cases.

Inheritance pattern Example Offspring odds from two carrier parents Who shows the trait
Autosomal dominant Huntington disease 50 percent affected One copy sufficient
Autosomal recessive Cystic fibrosis 25 percent affected Two copies required
X-linked recessive Hemophilia A 50 percent of sons affected Mostly males
X-linked dominant Fragile X syndrome 50 percent of children affected Both sexes, often milder in females
Mitochondrial Leber hereditary optic neuropathy All children of an affected mother Maternal line only
Polygenic Height, blood pressure Continuous distribution Threshold or gradient

For X-linked recessive conditions, the sex of the child changes the math. A carrier mother passes the variant to half her sons, who each have one X chromosome and therefore show the condition. Half her daughters become carriers. Fathers pass X-linked variants to all daughters and no sons. This asymmetry explains why hemophilia A appears in about 1 in 5,000 male births, per CDC figures, and is far rarer in females.

Epigenetics: inheritance beyond the DNA sequence

DNA sequence is not the whole story. Chemical marks attached to DNA and to the histone proteins that package it alter gene activity without changing the letters. The best-studied is DNA methylation, the addition of methyl groups to cytosine bases, which generally quiets gene expression. These marks are established in early development and maintained through cell division.

Two mechanisms make this heritable. First, mitotic inheritance: when a cell divides, maintenance enzymes copy methylation patterns onto the new DNA strand, so a liver cell gives rise to more liver cells with the same expression program. Second, transgenerational effects: some marks survive the reprogramming that strips most methylation from gametes and the early embryo. The scope of that survival is the subject of active research and is narrower than popular accounts suggest. Most epigenetic marks are erased between generations.

Imprinting is the clearest human example. For a small number of genes, expression depends on which parent supplied the copy. At the IGF2 locus, only the paternal copy is active; the maternal copy is silenced by methylation. Prader-Willi and Angelman syndromes arise from the same chromosomal region but produce different conditions depending on whether the deleted or silenced copy came from the father or the mother. That is a direct demonstration that parental origin, not just sequence, carries information.

A counterintuitive detail worth holding onto: identical twins share essentially the same DNA sequence, yet their methylation patterns diverge over a lifetime. Studies of twins show epigenetic differences accumulate with age and with divergent environments. So "same genes" does not mean "same biology," and the gap widens as people age.

Where the rules bend

The standard model has exceptions, and each one matters clinically.

New mutations: a child can carry a variant neither parent has. These arise in sperm or egg, or in the early embryo, at measurable rates. Every person carries an estimated 60 to 70 new single-letter variants not present in either parent. De novo structural changes: chunks of chromosome can be deleted, duplicated, or inverted during meiosis. These are common causes of developmental conditions and are not inherited from either parent. Mosaicism: a mutation arising after fertilization produces a person with two cell lineages. The fraction of affected cells, and which tissues they populate, determines the consequences. Uniparental disomy: a child inherits both copies of a chromosome from one parent and none from the other. This breaks imprinting rules and can cause conditions even with normal DNA content. Mitochondrial inheritance: mitochondrial DNA comes almost entirely from the egg, so mitochondrial variants pass through the maternal line. The exact transmission varies because mitochondria are distributed stochastically among eggs.

The professional caveat here is that "inherited" and "genetic" are not synonyms. A genetic condition can arise without any family history, and a family history can reflect shared environment as much as shared DNA. Genetic counselors treat a negative family history as a weak signal, not proof of low risk.

How environment modifies inherited potential

Most traits emerge from an interaction. The genotype sets a range; the environment determines where in that range the person lands. This is the norm, not the exception.

Height is the standard example. Twin studies and genome-wide analyses put height heritability at roughly 80 percent in well-nourished populations. That figure describes the fraction of variance explained by genetic differences across a population, not a fixed percentage for any one child. A child inheriting tall alleles in a setting of chronic malnutrition may not reach the predicted height. The heritability estimate also drops in populations where nutrition varies widely.

For complex diseases, the picture is more quantitative and less deterministic. Polygenic risk scores sum the small effects of many variants into a score that shifts probability, not certainty. A high score for type 2 diabetes raises risk but does not cause the condition in isolation. A low score does not exempt anyone. The CDC and major clinical societies emphasize that lifestyle factors remain the dominant modifiable drivers for most common chronic diseases.

The practical translation is that inherited risk is a prior probability. Actions and environment revise it over a lifetime. That framing is more accurate than either "genes decide everything" or "genes don't matter."

Factor Type of influence Modifiable?
Single-gene variant (e.g., CFTR) Strong, near-deterministic for the trait No
Polygenic risk score Shifts probability Partly, via risk factors
Epigenetic marks Alters gene expression Partly, via environment
Shared family environment Alters expression of risk Yes
New mutations Random, independent of parents No

Practical actions for parents and families

Inherited risk is actionable when the mechanisms are understood. The steps below are general guidance, not individualized medical advice.

Take a three-generation family history and bring it to a clinical visit. Record ages at diagnosis for relatives with cancer, heart disease, or diabetes. Early-onset disease in close relatives is the strongest family-history flag. Ask about carrier screening before pregnancy. Panel screening covers hundreds of recessive conditions, and most health plans cover it for people planning pregnancy. CDC-supported programs and professional genetics societies recommend discussing it with a clinician. Understand what a positive result means. A carrier result reports risk for offspring, not disease in the carrier. Couples where both partners carry the same recessive variant face a 25 percent chance per pregnancy and can discuss options with a genetic counselor. Treat polygenic scores as one input among several, not a verdict. Scores are most informative for a few conditions and least transferable across ancestry groups. Ask whether a score was validated in a population like yours. Distinguish risk from diagnosis. A family history of an autosomal dominant condition warrants discussion about testing. Testing without counseling can create more anxiety than clarity. Focus on modifiable levers for complex conditions. Blood pressure, blood glucose, tobacco use, and physical activity shift risk for most common diseases, whether or not inherited risk is elevated.

The practical ceiling is worth stating plainly. Genetic testing can clarify probabilities, but it cannot predict most individual outcomes. Counselors exist precisely because the numbers require interpretation.

Limits of what inheritance explains

Genetic inheritance is a robust framework, but it has documented boundaries.

Non-genetic inheritance exists and is real. Parents transmit microbiome components, cultural practices, and socioeconomic conditions alongside DNA. For some outcomes, such as child language exposure and educational attainment, shared environment explains more variance than sequence variants in current models.

Most common traits are not attributable to any single gene. Genome-wide association studies find that each common variant explains a tiny fraction of variance. The sum of discovered variants still leaves much heritability unexplained, a gap researchers call missing heritability. The gap likely reflects rare variants, gene-environment interactions, and epigenetic effects that current methods cannot yet measure well.

Prediction accuracy drops when models move across ancestry groups, because allele frequencies and linkage patterns differ. A score calibrated in one population may misestimate risk in another.

The honest bottom line: the rules for how traits pass from parents to offspring are well established for single-gene traits and probabilistic-but-real for everything complex. The mechanics explain a great deal. They do not yet explain all of it, and anyone who claims otherwise is overselling the science.

What parents transmit is both a fixed chemical inheritance and an adjustable one. The DNA sequence is largely set at fertilization. How it gets read unfolds over decades and responds to choices, environments, and care.