Some studies have claimed to show that handedness may begin as early as in the womb, suggesting a biological rather than a learned cause. Several studies have also shown a significantly increased frequency of lefthanders among those with central nervous system disorders (such as schizophrenia, epilepsy, mental retardation or learning disabilities) which may also be related to such environmental and biological factors.

A 2004 study of ultrasound observations by Peter Hepper et al showed that, as early as 9-10 weeks into gestation, fetuses can be observed moving their right arms significantly more than their left (around six times more). From about 12-15 weeks, around 9 out of 10 fetuses can be seen to be sucking their right thumb (a statistic that roughly matches the prevalence of right-handedness in the general population), which some have interpreted as evidence of a general predilection for right-handedness even before birth. Hepper followed up these findings, and confirmed that the right-handed thumb-suckers did indeed turn out to be right-handed, although a third of those who sucked their left thumbs in utero also turned out right-handed. One important aspect of these studies is that the thumb-sucking and arm-waving behaviour is being observed at a time before any asymmetry in the fetuses’ brains has developed (at this early stage, the brain is not even fully connect to the spinal cord and limbs).

Hepper has concluded from this that either handedness actually causes brain lateralization (and the not the other way round), or that there is no real link between handedness and the asymmetry of brains, or perhaps that the two phenomena occur side-by-side but reveal themselves at different times during development, all of which flies in the face of the generally accepted wisdom that handedness is caused by brain lateralization. Hepper has even suggested that maybe the roots of handedness develop, not in the brain itself, but in the brain stem or even in the spinal cord, and that perhaps it is just the side that develops first that becomes the dominant side.

But this is not the only way in which ultrasound has been linked with handedness. There may also be a weak association between the actual routine ultrasound screening of a fetus during pregnancy and left-handedness, as reported in the British Medical Journal in 1993, and in the Early Human Development journal in 1998. A Swedish study reported in the journal Epidemiology has also found evidence that males whose mothers underwent ultrasound scanning were significantly more likely to be left-handed than normal, and still more likely if the mothers received a second scan later in the pregnancy.

Most biological theories of handedness are so-called “pathological theories”, which essentially claim that left-handers are just brain-damaged right-handers. They start from the assumption that right-handedness is the natural norm, and that left-handedness is an aberration caused by one or more of various stressors or physical trauma, an assumption that many left-handers take issue with. Such theories revolve around the idea that, if the right half of the brain (which controls the left side of the body) is damaged during pregnancy or during the birth itself, then the default right-handedness of the baby remains unchanged. If, however, the left side of the brain is damaged due to some kind of trauma, then the brain’s natural plasticity will shift the responsibility for complex skilled movement to the right hemisphere of the brain, resulting in left-handedness.

Some studies have claimed that difficult or stressful births occur far more commonly among babies that grow up to be left-handed or ambidextrous. The Canadian psychologist Paul Bakan was among the first to suggest that left-handedness may be due to some extent to anoxia (oxygen deficiency), trauma or some other kind of damage to the left side of the brain during the birth process (although prima facie it is difficult to see how anoxia could affect one side of the brain and not the other).

Twins, both fraternal and identical, tend to show an elevated rate of left-handedness compared with singletons - almost double by some counts - and it has been argued that this could be due to the abnormal in utero conditions encountered in multiple birth pregnancies. The position of a fetus within the womb, and its position at birth, may also be related to handedness, and Fred Previc has hypothesized that an asymmetrical prenatal position may cause asymmetrical stimulation to the vestibular system of the inner ear (which governs spatial orientation and balance), which has also been linked to handedness.

Low birth weight has also been associated with left-handedness in some studies, and one possibility is that the brain development of such babies may have been interrupted at birth. Another study has suggested that left-handers are more likely to be born to first-time mothers or to older women, with the assumption again that elevated levels of birth stress are the main cause. A study reported in the New England Journal of Medicine suggests that mothers giving birth later on in life (as has become much more common in recent decades) are more likely to produce left-handed children. The study indicates that mothers between the ages of 30 and 35 have a 25% higher chance of producing a left-handed child.

It has even been claimed that such birth stress theories can also explain how left-handedness can run in families, at least to some extent. For example, if a mother has a smaller-than-average pelvis which causes a birth trauma resulting in a child’s left-handedness, then such a trait will likely be passed down to any girl child, who is in turn more likely to produce further brain-damaged left-handed progeny.

However, a detailed re-analysis of a 1958 study by the British National Child Development Study has showed no correlation between birth stress and left-handedness. Other studies have suggested that no more than perhaps 1-5% of left-handers might be created by in utero brain injuries, and a similar percentage due to pregnancy or birth complications.

Some studies indicate that even stressful life events of a non-physical nature during pregnancy can increase the likelihood of a left-handed or mixed-handed child. A 2003 Danish study reported in Developmental Medicine & Child Neurology suggests that mothers who experience multiple stressful events (e.g. the death of a loved one, a job loss, etc) during the third trimester of pregnancy are more than three times as likely to have a mixed-handed child than the general population. One possible cause suggested is the effects of cortisol, the body’s main stress hormone, which may interfere with the brains normal development.

One popular theory, developed by Norman Geschwind in the 1980s, revolves around the effects of elevated levels of testosterone at specific stages of pre-natal development. Such elevated levels may be caused by maternal stress, alcohol intake, excess body fat, physical activity or a number of other factors. Geschwind had already noted a connection between auto-immune diseases and high testosterone levels, and showed how testosterone can inhibit the development of the thymus gland (which helps to regulate the immune system). He then noted the widely reported (although still unproven) link between left-handedness and auto-immune diseases, and from there it was a short hop to associate high levels of testosterone with left-handedness. Geschwind also claimed that his testosterone theory accounts, at least to some extent, for the increased likelihood of left-handedness among males compared to females,

However, notwithstanding more recent data which might seem to support it (such as a 2011 French study which found higher than average testosterone levels in male left-handers), the theory has been largely discredited by several other studies in the 1990s. Although a popular speculation at the time, it remains far from proven that left-handers are more susceptible to immune disorders (see the section on Handedness and Health for more discussion of this). The theory also breaks down when considering twins who exhibit different handedness. It does not explain how testosterone can affect the brain hemispheres differently anyway, nor why, if testosterone is the major cause, the male-female gender gap in left-handedness (about 13% of males compared to 11% of females in the developed West) is not more pronounced than it actually is. Indeed, it has been argued that the gender discrepancy can be convincingly explained by cultural variables (men are apparently more likely to report extreme categories in general than women, who tend to be more moderate in their claims; women are more likely to respond to pressure to conform to norms than men, and so are more likely to switch to right-handedness in childhood; etc).

There is also some inconclusive evidence that other hormones may affect handedness. A study endorsed by the US Centers for Disease Control has suggested that pre-natal exposure to estrogen may influence left-handedness. Another study claims, more specifically, that individuals exposed prenatally to the synthetic estrogen-based drug diethylstilbestrol appear to be more likely to be left-handed.

Although very popular a few decades ago, these kinds of pathological theories of handedness theories are notoriously difficult to prove with any conviction, and most are considered improbable today, or at best contributing rather than causal factors. It is generally agreed that quite a significant brain injury in utero would be needed to cause a switch in handedness, an injury which would likely disrupt cognition in much more dramatic ways. Assuming that there is a 50-50 chance of any trauma occurring on either side of the brain, there would presumably also be a (smaller, but still significant) proportion of "pathological right-handers" born each year. Also, the theories do not explain the persistence of left-handedness in an era of much improved medicine and childbirth technology, nor why humans should be right-handed by default in the first place.

Any link between retardation in general and left-handedness appears to be weak at best. Various studies over the years have purported to show that left-handers are more likely to develop developmental disabilities, immune disorders, autism, dyslexia, epilepsy, hyperactivity, schizophrenia and even alcoholism and criminality, but many of these do not stand up to scrutiny (see the section on Handedness and Health). A major meta-analysis by Chris MacManus and Clare Porac among others has concluded that left-handers are in fact no more likely to suffer from auto-immune disorders, or indeed health problems in general, than right-handers. In particular, it should be noted that, when looking at severely intellectually-challenged individuals, the degree of sensitivity between right- and left-handed performance is much smaller and more difficult to measure, and the whole concept of handedness may become moot if they lack the specialized skills needed to determine hand preference in the first place.

Just for the record, a couple of other (discredited) biological theories should be mentioned. One old idea, sometimes called the Vanishing Twin Theory, is that every left-hander was once a twin whose right-handed mate did not make it through gestation. Another rather simplistic and probably unprovable theory is that those fetuses whose left arms face outwards in the womb tend to become left-handed as they have more space and opportunity to develop a stronger and more dominant left side while still in utero.