Looks just like daddy: Investigating perceptions of fetal resemblance

– By Carlota Batres

When it comes to children, women can be certain that a child is theirs, but men cannot. In other words, there is always some chance that another man fertilized the woman’s egg. The estimate of children who have genetic fathers different from their alleged fathers is between 1%-3%.

One of the sources of information that men use to determine the likelihood of being the genetic father is perceptions of the child’s physical similarity. Research has found that fathers who perceive that their children resemble them invest more in those children, for example by spending more time with them and by having more supportive/engaged parenting.

In our study, we aimed to examine whether mothers having an ultrasound would claim that the fetus looks more like the father. Even with such limited visual information, we predicted that mothers would be more likely to say that the fetus resembled the father in an effort to assure paternity.

In addition to perceptions of the child’s resemblance, men also use information about their partner’s faithfulness and trustworthiness to determine how likely they are to be the biological father. For instance, one study found that men’s perceptions of their partner’s fidelity predicted their parental investment. Marriage is one way in which couples pledged faithfulness to each other. Indeed, while infidelity takes place in married couples, the incidence of cheating is much lower than it is in unmarried couples. Therefore, for our study, we also aimed to examine differences between parents who were married and parents who were not. We predicted that among unmarried parents, mothers would be even more likely to say that the fetus resembled the father.

Participants were recruited at an obstetrics and gynecology clinic, mostly during their routine 20-week pregnancy ultrasound. If they decided to participate, they were given a questionnaire and then called in for their ultrasound. Once the fetus was visible in the 2D imaging, if the biological mother and the biological father were in the ultrasound room (as per the participants’ questionnaire responses), the ultrasound technician would ask them “Who do you each think the baby looks like?” and would record their answers.

We found that fathers were just as likely to say the fetus resembled themselves (49%) as they were to say that the fetus resembled the mother (51%). However, mothers were significantly more likely to say that the fetus resembled the father than to say that the fetus resembled themselves. In fact, 26% said that the fetus resembled themselves and 74% said that the fetus resembled the father. Moreover, mothers who were not married were even more likely to say that the fetus resembled the father than married mothers. More specifically, 93% of unmarried women said that the fetus resembled the father.

Evidence from our study supports the hypothesis that mothers should be more likely to say that the fetus resembles the father. If the fetuses actually did resemble the fathers, the fathers should also be more likely to say that the fetuses look like them. However, we did not find this to be the case. Rather, we only saw this effect with mothers, suggesting that the unconscious bias in mothers’ observations about similarity to the fathers does not reflect actual resemblance. This is supported by other research which has found that when unrelated participants were asked to match photographs of parents to newborns, they were more accurate in matching the mothers than the fathers.

In conclusion, our study provides evidence that even with extremely limited visual information, mothers have a bias toward claiming paternal similarity. This bias is likely an adaptation that helps the mothers secure investment in their unborn children since fathers who believe that their children resemble them invest more in those children.

Read the original article: Batres, C., Mullen, A., Krofl, S., & Trainor, L. (2025). Investigating perceptions of fetal resemblance. Evolution and Human Behavior, 46(2), 106670.