Two women cannot conceive a child together using their own bodies alone. Human reproduction requires the fusion of a sperm cell and an egg cell, and no natural biological process allows one woman to produce sperm or another to fertilize an egg without it. But that straightforward answer is increasingly incomplete. Over the past two decades, laboratory breakthroughs in mice have produced live, healthy offspring from two mothers with no genetic father involved. Researchers have also generated early-stage artificial sperm from female stem cells. None of this is ready for human use, and serious biological hurdles remain, but the science has advanced far enough that the question has shifted from “is it possible in theory” to “how many obstacles are left.”
Why Natural Conception Requires Sperm
Fertilization is a multistep process that culminates in the fusion of a sperm and egg, initiating the development of a new organism.1PubMed. The Fertilization Enigma: How Sperm and Egg Fuse A mammalian egg sits inside two protective layers, and sperm must physically penetrate both, bind to the egg’s surface, and fuse with it to deliver its genetic payload.2PubMed. Penetration, adhesion, and fusion in mammalian sperm-egg interaction Without sperm, the egg has no way to receive a second set of chromosomes and begin dividing normally.
But the need for sperm goes deeper than just delivering DNA. In humans, the sperm also provides a structure called the centrosome, which organizes the first cell divisions of the embryo. The human egg’s own centrosome is inactive, so the sperm’s centrosome is what actually drives the early cell-division machinery.3PubMed. Mitosis in the human embryo: the vital role of the sperm centrosome (centriole) This means that even if you could somehow combine two eggs’ worth of DNA, you would still face the problem of getting the resulting cell to divide properly. Unlike mice, where the egg retains functional centrosomes, human centrioles are paternally derived.4PubMed. Centrioles in the beginning of human development
The Imprinting Problem
Even setting aside the centrosome, there is a more fundamental barrier: genomic imprinting. Not all genes behave the same way depending on which parent they came from. During the formation of eggs and sperm, certain genes get chemically tagged so that only the mother’s copy or only the father’s copy is active in the offspring. This tagging means a developing embryo needs one maternal and one paternal genome to function. An embryo with two maternal genomes has the wrong pattern of active and silent genes, and in mammals, it fails to develop normally.5PubMed. Genomic imprinting is a barrier to parthenogenesis in mammals
This is what makes mammals different from many reptiles and insects. Some lizard species can reproduce through parthenogenesis, where females produce offspring from unfertilized eggs. Researchers studying whiptail lizards found that females housed with males still sometimes produced parthenogenetic offspring alongside sexually fertilized ones, suggesting the capacity is built into their biology rather than triggered by male absence.6eLife. Post-meiotic mechanism of facultative parthenogenesis in gonochoristic whiptail lizard species Mammals locked themselves out of this reproductive shortcut through imprinting, which evolved to regulate growth and resource allocation between mother and offspring.
What Same-Sex Female Couples Can Do Right Now
Today, the practical answer for two women who want to start a family involves donor sperm. But reproductive medicine has developed ways to let both partners participate biologically. In reciprocal IVF, one partner’s eggs are retrieved and fertilized with donor sperm, and the resulting embryo is transferred into the other partner’s uterus. This means one woman is the genetic mother and the other is the gestational mother. A systematic review covering more than 1,400 reciprocal IVF cycles found that pregnancy and birth outcomes were comparable to standard IVF.7PubMed Central. A Systematic Review of Reproductive Technologies for Shared Conception in Same Sex Female Couples
A large study published in JAMA confirmed these findings. After adjusting for relevant factors, the clinical pregnancy rate for reciprocal IVF was about 64% compared to 57% for standard autologous IVF, with live birth rates of roughly 52% versus 48%. The differences were not statistically meaningful. There was a slightly higher rate of singleton preterm birth among reciprocal cycles, about 15% compared to 11%, though this too fell short of clear statistical significance.8JAMA. Outcomes of Same-Sex Female Couples Pursuing Autologous and Reciprocal In Vitro Fertilization Reciprocal IVF is widely available and gives both women a physical role in the pregnancy, but it still requires a male genetic contribution through donor sperm.
Bimaternal Mice and the Breakthroughs That Made Them Possible
The most dramatic proof that two-mother offspring can survive and thrive in mammals comes from mouse experiments. In 2007, researchers created bimaternal mice by combining genetic material from two types of eggs: fully grown oocytes and immature non-growing oocytes from donor mice carrying specific genetic deletions. By knocking out the imprinting control regions at the H19 and Dlk1-Dio3 gene clusters, they corrected the growth-gene imbalance that normally kills bimaternal embryos. These mice developed at a success rate equivalent to conventional IVF, grew into fertile adults, and lived normal lifespans.9PubMed. High-frequency generation of viable mice from engineered bi-maternal embryos
A 2018 study pushed this further using CRISPR gene-editing tools. Researchers used haploid embryonic stem cells with deletions in three imprinted regions, injected them into normal eggs, and obtained healthy, fertile bimaternal mice. The same team also attempted bipaternal mice (two fathers, no mother), which required deleting seven imprinted regions. Those pups were born alive but died shortly after, illustrating that the maternal contribution is even harder to replace than the paternal one.10Cell Stem Cell. Generation of Bimaternal and Bipaternal Mice from Hypomethylated Haploid ESCs with Imprinted Region Deletions Another group of researchers described the work as effectively “CRISPRing out the barriers for unisexual reproduction.”11Cell Stem Cell. Mice from Same-Sex Parents: CRISPRing Out the Barriers for Unisexual Reproduction
Then in 2022, another team reported something even more striking: live mouse offspring from single unfertilized oocytes, achieved by rewriting the DNA methylation at seven imprinting control regions.12PubMed Central. Viable offspring derived from single unfertilized mammalian oocytes These results demonstrate that the imprinting barrier, while real, is not insurmountable. It is an engineering challenge, not an absolute biological law.
Making Sperm from Female Cells
A completely different approach skips the imprinting workaround entirely and instead tries to manufacture sperm from a woman’s own cells. This falls under a field called in vitro gametogenesis, or IVG. Researchers have already coaxed human stem cells into germ cell-like cells at various stages of development.13PubMed Central. Artificial gametes from stem cells A systematic review noted that artificial sperm has been generated from female induced pluripotent stem cells in human cell lines.14Human Reproduction Update. Artificial gametes: a systematic review of biological progress towards clinical application
In mice, the technology has gone much further. Mouse stem cells can be induced into functional eggs and sperm, while human stem cells have so far been pushed only to early-stage eggs and immature sperm precursors.15PubMed. Mammalian in vitro gametogenesis One Japanese experiment made headlines by reprogramming skin cells from an adult male mouse into stem cells, growing them until some spontaneously lost their Y chromosome, then driving those cells to produce functional eggs. Those eggs were fertilized with sperm from a different male mouse, and the resulting embryos produced live pups, meaning two males had genetic offspring together.16PubMed Central. Pluripotent Stem Cell-Derived In Vitro Gametogenesis and Synthetic Embryos—It Is Never Too Early for an Ethical Debate The reverse scenario, creating sperm from female cells, faces a unique problem.
The Y Chromosome Gap
Women carry two X chromosomes and no Y chromosome. The Y chromosome contains genes in a region called the azoospermia factor (AZF) region that are essential for producing sperm.17PubMed Central. Y chromosome azoospermia factor region microdeletions and transmission characteristics in azoospermic and severe oligozoospermic patients The most frequently studied of these, the DAZ gene, is directly involved in germ cell development, and deletions in it are among the most common genetic causes of male infertility.18PubMed Central. Azoospermia Factor C Subregion of the Y Chromosome
For two women, this means any sperm-creation approach would need to either introduce the critical Y-chromosome genes artificially or find a way to complete sperm development without them. There is no proven method to do either in human cells. Even if the DNA-packaging steps could be replicated, the resulting sperm would need to carry either an X or a Y chromosome to determine the offspring’s sex. Without a Y chromosome anywhere in the picture, all offspring from two genetic females would necessarily be female themselves. That would not be a safety concern per se, but it is a biological constraint worth understanding.
A recent mouse study explored a related approach: creating artificial oocytes by transferring nuclei from somatic cells into enucleated eggs. Using endometrial stromal cells, researchers achieved blastocyst formation and transferred 42 blastocysts, yielding three healthy pups that grew to adulthood and reproduced normally. All three showed biparental inheritance, confirming the somatic cell genome had combined with sperm DNA successfully.19PubMed. Artificial oocytes through haploidization of mouse endometrial stromal cells and bone marrow-derived mesenchymal stem cells This shows that somatic cells can be repurposed for reproduction in mice, though the efficiency remains low and the jump to humans is enormous.
Mitochondrial Replacement as a Partial Solution
A more immediately realistic technology would not give two women a fully shared genetic child but would let both contribute genetically in a limited way. Mitochondrial replacement techniques, originally developed to prevent the transmission of mitochondrial diseases, work by placing the nuclear DNA from one woman’s egg into the enucleated egg of another woman, preserving the second woman’s mitochondrial DNA. The resulting child would carry nuclear DNA from one mother (and a sperm donor) plus mitochondrial DNA from the other mother.
Researchers have argued that lesbian couples should be allowed access to this technology. The reasoning is that the moral case for offering mitochondrial replacement to women at risk of transmitting mitochondrial disease, which centers on reproductive freedom and the absence of harm to the child, applies equally to same-sex couples who wish for a genetic link to their offspring.20PubMed Central. Lesbian motherhood and mitochondrial replacement techniques: reproductive freedom and genetic kinship This is far from a full genetic merging of two women’s DNA, since mitochondrial DNA represents less than 0.1% of the genome, but it is a way to make both partners biologically present in the child’s cells.
Where Transgender Reproduction Fits In
A transgender woman (assigned male at birth) who has preserved sperm or testicular tissue before or during her transition can use that material to conceive with a cisgender female partner through standard assisted reproduction.21PubMed Central. Fertility preservation options for transgender individuals In this scenario, two women, one transgender and one cisgender, can have a genetically shared child today using existing IVF techniques. The legal and social recognition of this varies by jurisdiction, but the biology is straightforward: one partner provides sperm and the other provides the egg and gestational capacity, or vice versa with the help of donor eggs.
Uterine transplantation has also raised theoretical questions about whether transgender women could carry pregnancies. Uterus transplants from both living and deceased donors have produced live births in cisgender women born without a uterus. One program reported a live birth rate of about 79% per technically successful transplant, with 12 babies born across 14 transplants.22PubMed. Twelve Live Births After Uterus Transplantation in the Dallas UtErus Transplant Study The first live birth from a deceased-donor uterus transplant was reported in a patient with a congenital absence of the uterus, establishing proof of concept for the approach.23The Lancet. Livebirth after uterine transplantation from a deceased donor Whether this could work in a natal male body remains entirely theoretical. One anatomical study found that internal iliac artery dimensions are similar between sexes, which is relevant because those arteries supply the transplanted uterus, but there are countless other anatomical and hormonal challenges that have never been tested.24BJOG: An International Journal of Obstetrics and Gynaecology. Uterine transplantation in transgender women
Safety Concerns with Manipulated Embryos
Even if the technical barriers to two-mother conception were solved tomorrow, the safety questions would take years to answer. Assisted reproduction already raises epigenetic concerns. The period around conception is critical for establishing the chemical tags on DNA that guide development, and lab culture conditions can disrupt this process. The worry is not just about rare imprinting disorders but about subtle shifts in the epigenome that could influence disease risk later in life.25Fertility and Sterility. Epigenetic disturbances in in vitro cultured gametes and embryos: implications for human assisted reproduction
The techniques needed for bimaternal reproduction, including CRISPR editing of imprinted regions, artificial gametogenesis from stem cells, and nuclear transfer, each add layers of epigenetic manipulation far beyond what standard IVF involves. In the mouse experiments that produced bimaternal pups, the success rate was low and many embryos failed at various stages, suggesting the reprogramming was incomplete in most attempts. Scaling this to humans, where the consequences of getting it wrong affect a person who lives for decades, is a fundamentally different problem from producing a research mouse that lives for two years.
The Regulatory and Ethical Landscape
No country currently permits the clinical use of IVG-derived gametes for human reproduction. Regulatory bodies are beginning to consider how they would evaluate such technologies if they reached the point of clinical readiness. The ethical questions include safety, the destruction of embryos during research, equitable access, and the possibility of creating gametes from someone without their consent.26Fertility and Sterility. Ethical and legal implications of in vitro gametogenesis and germline editing—current status
An analysis of how Singapore might regulate these technologies offers a window into likely global patterns. IVG is expected to receive narrow approval for treating primary infertility and age-related egg decline, and potentially for facilitating transgender parenthood and preventing genetic diseases. Same-sex parenthood applications, however, along with solo parenthood and genetic enhancement, would likely remain banned in that jurisdiction.27PubMed. Ethical, Legal and Social Issues in Utilizing In Vitro Gametogenesis (IVG) and Stem Cell-Based Embryo Models (SCBEMs) for Human Reproduction in Singapore Even in more permissive regulatory environments, the definition of biological parenthood itself is becoming contested. A scoping review found no consensus on what “biological parent” means when applied to these technologies, which raises problems for legal frameworks that use biological parentage to assign rights and responsibilities.28SpringerLink / J Bioeth Inq. Biological Parenthood in Relation to Future Artificial Reproductive Technologies: A Scoping Review of Definitions
What We Know About Children in Same-Sex-Parent Families
While the technology to create a child from two women’s DNA alone does not yet exist for humans, plenty of children are already growing up with two mothers through donor conception and adoption. A longitudinal study following adoptive families with school-age children found no differences in child adjustment, parenting quality, couple functioning, or family dynamics based on whether the parents were lesbian, gay, or heterosexual. The best predictors of how children fared were earlier child adjustment and parenting stress, not parental sexual orientation.29PubMed. Does parental sexual orientation matter? A longitudinal follow-up of adoptive families with school-age children This matters for the broader conversation because it addresses one of the common anxieties people voice when new reproductive technologies are proposed for same-sex couples. The evidence consistently shows that the children do fine.
How Far Off Is the Real Thing
The honest answer is that nobody knows whether two cisgender women will ever be able to have a child who shares both their genomes and no one else’s. The mouse experiments are genuine scientific achievements, not speculative thought experiments. Fertile, healthy bimaternal mice exist. But the distance between mice and humans in reproductive biology is unusually large. The centrosome issue alone, which does not exist in mice, adds a layer of complexity that has not been addressed in any bimaternal reproduction study. Human imprinting is more complex than mouse imprinting, with more imprinted genes and less certainty about which ones would need to be edited. And the ethical requirements for human trials, where a child’s lifelong health is at stake, mean that decades of animal safety data would likely be needed before anyone attempts it clinically.
IVG researchers working with human stem cells have produced germ cell-like cells but have not yet created fully functional human eggs or sperm from them. The gap between a cell that looks like a gamete and one that can reliably produce a healthy embryo is enormous. That said, the pace of progress has surprised even people in the field. The cloning of Dolly the sheep in 1996, which showed that adult cell nuclei could be reprogrammed to an embryonic state, was considered a decade ahead of its time.30PubMed Central. Somatic cell nuclear transfer: origins, the present position and future opportunities Within ten years, the discovery of induced pluripotent stem cells made reprogramming routine. Reproductive biology has a pattern of advancing in sudden leaps after long plateaus, and the tools available now, particularly CRISPR and improved stem cell culture, are more powerful than anything previous generations of researchers had access to.