Two men cannot yet have a child who is genetically related to both of them, but the science is closer than most people realize. In 2023, a Japanese research team created baby mice whose genetic material came entirely from two male parents, using stem cells from one male reprogrammed into functional egg cells. Since then, multiple independent groups have produced adult bipaternal mice through different techniques. The gap between a mouse proof-of-concept and a human clinical option remains enormous, but it is no longer a question of whether mammalian biology allows it in principle.
Making Egg Cells From Male Cells
The most direct route to two-father offspring involves turning one father’s cells into egg cells. In the landmark 2023 experiment, researchers at Osaka University took skin cells from a male mouse, reprogrammed them into induced pluripotent stem cells, and then guided those cells through a process that deleted the Y chromosome and duplicated the X. The resulting cells, now carrying two X chromosomes, were coaxed into becoming mature oocytes in a dish. When fertilized with sperm from a second male mouse, those lab-made eggs produced viable offspring.
1PubMed. Generation of functional oocytes from male mice in vitroAn earlier proof of the same idea used a slightly different strategy. Male mouse cells were reprogrammed into pluripotent stem cells, and researchers waited for some of those cells to spontaneously lose their Y chromosome, a known quirk of stem cell culture. The resulting XO cells (one X, no Y) were used to create female chimeric mice, which were then bred with a second male. The pups carried genetic contributions from both fathers.
2PubMed Central. Generation of viable male and female mice from two fathersThese experiments proved that male mammalian cells can, under the right conditions, produce eggs capable of generating healthy offspring. But they also revealed how roundabout the process is. Losing or replacing a sex chromosome, coaxing cells through dozens of developmental stages in a dish, and then confirming that the resulting eggs are chromosomally normal at every step all require meticulous control. Even in mice, only a fraction of the engineered eggs develop into healthy pups.
The Imprinting Barrier
Even if you could make a perfect egg from one man’s cells and fertilize it with another man’s sperm, you would still face the problem of genomic imprinting. In normal reproduction, certain genes are chemically tagged so that only the copy inherited from the mother or only the copy inherited from the father is active. When both copies come from the same sex, this tagging goes haywire.
Classic experiments in mice demonstrated this decades ago. Embryos engineered to carry two sets of paternal chromosomes and no maternal ones (called androgenetic embryos) develop poorly: the embryo itself barely forms, while the placenta-supporting tissues are somewhat better. The reverse, embryos with two maternal genomes, show the opposite pattern: the embryo starts growing but placental tissue fails.
3Cell Press (Current Biology). Genomic imprintingNeither type survives to birth without intervention. This is the biological enforcement mechanism that makes same-sex reproduction difficult in mammals. The genome essentially expects one set of instructions from a mother and a different set from a father. Getting around that expectation requires rewriting the chemical marks on specific stretches of DNA.
Rewriting the Marks in Bipaternal Mice
Rather than converting male cells into eggs, some research groups have attacked the imprinting problem head-on by editing the epigenetic tags that prevent two-father embryos from surviving. In 2025, a team published results showing that frameshift mutations, gene deletions, and regulatory edits at 20 imprinted loci allowed bipaternal mouse embryos to develop into adults. The survival rate was low, but the animals that made it were fully grown.
4Cell Stem Cell. Generation of Viable Bipaternal Mice from Haploid Stem Cells with Imprinting ModificationsA separate group used CRISPR-based epigenome engineering to take a more targeted approach. Instead of permanently altering gene sequences, they edited the methylation patterns on seven imprinting control regions. Two sperm cells were injected into an enucleated egg (one that had its own nucleus removed), and guide RNAs selectively reset the chemical marks on one sperm’s contribution to mimic a maternal pattern. The resulting androgenetic mice reached adulthood and were fertile.
5PubMed Central. Fertile androgenetic mice generated by targeted epigenetic editing of imprinting control regionsFollow-up work examined what happens to these bipaternal mice in more detail. Even with 20 imprinting modifications, placental abnormalities persisted. To work around this, researchers used a technique called tetraploid complementation, where normal cells form the placenta while the bipaternal cells contribute only to the embryo itself. This separated two distinct problems: the embryo’s own development and the placenta’s function.
6PubMed. Engineered bipaternal mice reveal the consequences of life without a maternal genomic contributionCollectively, these studies confirmed that imprinting abnormalities are the primary barrier to same-sex mammalian reproduction, not some deeper incompatibility in the genome. The barrier is formidable, but it is a specific engineering challenge rather than a fundamental impossibility.
Why Human Applications Are Still Distant
Mouse results tend to be reported as if the human version is around the corner. It is not, and the reasons go beyond the usual caveats about animal research. The state of the science in humans is considerably further back than in mice. While mouse stem cells have been coaxed all the way to functional eggs and sperm, human stem cells have only been guided to much earlier developmental stages: early oocyte-like cells and prospermatogonia, which are precursors to mature sperm rather than the real thing.
7Science. Mammalian in vitro gametogenesisThe practical gap is significant. In vitro gametogenesis, the umbrella term for making gametes in a lab from stem cells, requires guiding cells through an intricate developmental program that takes years in a human body. Mouse eggs mature much faster and in a much smaller organ. Scaling the culture conditions, growth factor cocktails, and quality-control checkpoints from one species to the other is not a minor adjustment; it is a different engineering project.
There is also the issue of genetic fidelity. Lab-grown cells accumulate mutations at a considerably higher rate than natural germ cells because the body’s specialized reproductive tissue has quality-control mechanisms that do not exist in a petri dish. One analysis noted that the mutation rate in somatic cells is roughly tenfold higher than in germ cells, meaning that any gamete derived from reprogrammed skin or blood cells starts at a disadvantage. Failures in imprint erasure, imprint resetting, or the cell-division process of meiosis can introduce additional errors.
8Stem Cells Translational Medicine. Pluripotent Stem Cell-Derived In Vitro Gametogenesis and Synthetic Embryos—It Is Never Too Early for an Ethical DebateNo responsible regulator would approve clinical use of lab-derived human gametes without extensive safety data, and generating that data will take years of further animal work, followed by cautious steps in human cells that never involve creating a pregnancy. Any child born from this technology would have to be the end product of a safety pipeline that does not yet exist.
Who Carries the Pregnancy
Even if the genetic puzzle were solved tomorrow, two men still need a way to gestate a pregnancy. Today, that means a gestational surrogate: a woman who carries the embryo. For male couples who use surrogacy now, only one father contributes sperm; the egg comes from a donor. The hypothetical future scenario where both fathers contribute genetically would still require either a surrogate or some alternative gestational technology.
Artificial womb technology gets a lot of attention in this context, but its current state is far more limited than popular articles suggest. The most successful animal experiments have maintained extremely premature lamb fetuses in fluid-filled chambers for up to about four weeks, supporting gas exchange and fetal circulation. Most platforms achieve only about a week of support. No human trials have been conducted, and the technology is designed to help extremely premature infants born at 22 to 24 weeks, not to replace the full 40 weeks of gestation.
9PubMed Central. Artificial Womb Technology: A Systematic Review of Preclinical Evidence and Implications for Neonatal Viability and Intensive CareResearchers have also developed early embryo culture systems, uterine organoids, and engineered endometrial models, but these reproduce only stage-specific functions of the uterus or placenta. Supporting the full dynamic process of human gestation from fertilization to birth remains well beyond current capabilities.
10Chemical Engineering Journal. Artificial womb technology: Progress, challenges, and future directions toward extrauterine gestationAnother speculative possibility is uterine transplantation. Uterus transplants have already resulted in live births in cisgender women, and feasibility analyses have explored the idea of transplanting a uterus into a transgender woman’s body. One study found that the internal blood vessels needed for the surgical connection are similarly sized in male and female bodies, with only a modest caliber difference of about 1.6 mm that could be managed by adjusting where the surgical join is made.
11PubMed Central. Uterine transplantation in transgender womenWhether a transplanted uterus in a male-bodied person could sustain a full pregnancy remains unknown. The hormonal environment, pelvic anatomy, and immune considerations are all uncharted territory. No such transplant has been attempted in a cisgender man or a transgender woman to date.
How Children of Gay Fathers Are Doing
While the genetic science is still catching up, thousands of children are already growing up with two fathers through surrogacy, adoption, and other arrangements. A concern sometimes raised is whether children need a mother for healthy development. The research on this question is reassuringly clear. A meta-analysis examining psychological adjustment in children of planned gay father families found that all children scored below the clinical threshold for behavioral and emotional problems. In fact, children of gay fathers showed slightly better adjustment than children of heterosexual parents across the studies analyzed.
12Sexuality Research and Social Policy. Child Psychological Adjustment in Planned Gay Father Families: A Meta-analysisThese findings are consistent with decades of research on same-sex parent families more broadly. The consistent conclusion is that parenting quality, economic stability, and social support matter far more for child outcomes than the sex of the parents. If and when two-father genetic parenthood becomes possible, the child-wellbeing question is likely to center on the safety of the reproductive technology itself rather than on the family structure.
Two-Father Offspring in the Animal Kingdom
The idea of offspring produced entirely from male genetic material sounds like pure science fiction, but it occurs naturally in at least one vertebrate species. In a freshwater fish complex called Squalius alburnoides, researchers documented the first known case of spontaneous androgenesis in a vertebrate: an offspring whose nuclear DNA came entirely from its father, with the mother’s genetic contribution discarded during fertilization.
13PubMed Central. First empirical evidence of naturally occurring androgenesis in vertebratesThis is not exactly two-father reproduction; it is one father cloning himself through an egg. But it shows that the biological machinery of an egg cell can, under unusual circumstances, be hijacked to support development using only paternal DNA. The phenomenon is rare even in fish: the observed frequency of androgenetic offspring in vertebrate studies has ranged from about one to five percent. Still, the fact that it happens at all in nature shows that the egg’s developmental machinery is not inherently dependent on maternal chromosomes for every step of embryogenesis. Mammals layer on extra restrictions through imprinting, which is why the process does not happen spontaneously in any mammalian species.
14Royal Society Open Science. First empirical evidence of naturally occurring androgenesis in vertebratesThe Ethics Conversation Already Underway
Ethicists are not waiting for the science to be ready. The prospect of in vitro gametogenesis has already sparked serious debate about who should have access, what safeguards are needed, and whether the technology could create pressures nobody intended. Among the concerns: the cost could make it a luxury available only to wealthy prospective parents, deepening existing inequalities in reproductive medicine. The burden of egg retrieval could be eliminated for women, which is broadly positive, but the same technology could also pressure women to delay parenthood further in the belief that lab-made eggs will always be available as a backup.
8Stem Cells Translational Medicine. Pluripotent Stem Cell-Derived In Vitro Gametogenesis and Synthetic Embryos—It Is Never Too Early for an Ethical DebateFor male couples specifically, the current reality of surrogacy already involves navigating complex emotional, legal, and market dynamics. Research on gamete selection in gay male surrogacy has found that decisions about which partner’s sperm to use, which egg donor to select, and how to structure the relationship with the surrogate are shaped by a tangle of biomedical constraints, legal frameworks, and personal values. Efforts to protect children from discrimination sometimes inadvertently reproduce racialized and gendered hierarchies embedded in the broader reproductive marketplace.
15PubMed. Gamete selection in gay male surrogacy: Procreative labor beyond gestationOne question that in vitro gametogenesis would resolve for many gay couples is the asymmetry of genetic parenthood. When only one father can contribute sperm and the other has no genetic link to the child, the legal and emotional dynamics within the family can be uneven. Two-father genetic parenthood could eliminate that asymmetry, but it would introduce new questions about mitochondrial inheritance. Any lab-made egg cell would still need a donor egg’s cytoplasm, including its mitochondria, meaning a third person’s genetic material would be present in every cell of the child’s body. The ethical debate about whether an egg donor for mitochondrial replacement has parental responsibilities is already active in the context of mitochondrial donation for disease prevention.
16PubMed Central. Does egg donation for mitochondrial replacement techniques generate parental responsibilities?The Mitochondrial Wrinkle
Even in the most optimistic future scenario, a child of two men would not be 100 percent genetically derived from those two men. Mitochondria, the energy-producing structures inside every cell, carry their own small genome and are inherited almost exclusively through the egg. If a lab-made oocyte is constructed inside a donor egg’s shell, the child inherits the donor’s mitochondrial DNA alongside the nuclear DNA from both fathers.
This is a tiny fraction of the total genome, only 37 genes compared to roughly 20,000 in the nucleus, but it is inherited material from a person who is neither legal parent. The same situation already exists in children born through mitochondrial replacement therapy, a technique approved in the United Kingdom to prevent mitochondrial diseases. Whether that small genetic contribution creates any form of parental relationship, moral or legal, is an open question that bioethicists are actively debating. For two-father reproduction, the mitochondrial donor would be functionally invisible in terms of the child’s traits and appearance but technically present in the child’s genetic makeup.
What a Realistic Timeline Looks Like
Predictions about timelines in reproductive biology have a poor track record, but a rough sense of the roadmap is possible. The immediate next steps involve getting human stem cells past the early germ cell stage and into functional, mature gametes in a dish. That milestone has not been reached and could take a decade of incremental progress. After that, extensive safety testing of lab-derived gametes in non-human primates would be needed before any regulatory body would consider human trials.
The imprinting-editing approach faces its own bottleneck. Editing seven to twenty imprinting control regions with enough precision to produce a healthy embryo in mice is already difficult; doing it reliably in human cells, where the stakes of error are immeasurably higher, will require a level of epigenetic engineering that does not yet exist. Each species has its own imprinting landscape, so the specific edits that work in mice will not simply transfer to human embryos.
If everything went well, the most optimistic informal estimates from researchers in the field place a first demonstration of human two-father embryos (not necessarily a pregnancy, just a viable embryo in a dish) somewhere in the 2030s, with clinical availability for actual pregnancies unlikely before mid-century. These timelines could compress with unexpected breakthroughs or stretch with unexpected safety problems. The one thing that is clear is that the science has moved out of the realm of pure thought experiment and into the realm of engineering problems with identifiable, if difficult, solutions.