Here is the content of @Coloma link:
Genetics and Evolution – Theories About The Evolution of Handedness
One characteristic that distinguishes humans from other primates is the fact that the overwhelming majority (90%) are right-handed. With the exception of parrots, which show a similar population bias for right-footedness, other animals appear to have more balanced proportions of hand or foot preferences. In “The Genetics and Evolution of Handedness” (Psychological Review, vol. 104, no. 4, 1997), Michael Corballis proposed that at some time during human evolution a mutation produced a “dextral” allele (D), strongly predisposing humans to right-handedness. Corballis speculated that the trait was additive—that is, it was also influenced by other genes that supported the bias toward right-handedness—and that the D allele, with a slight heterozygotic advantage, was probably spread and generalized to the population quite quickly. This inheritance would explain the relatively unchanged proportions of right- and left-handed people in the population. Corballis hypothesized that this unique human allele may have arisen in Africa between 150,000 and 200,000 years ago.
Gregory Jones and Maryanne Martin presented a somewhat different mechanism of the inheritance of handedness in “A Note on Corballis (1997) and the Genetics and Evolution of Handedness: Developing a Unified Distributional Model from the Sex-Chromosomes Gene Hypothesis” (Psychological Review, vol. 107, no. 1, 2000). The researchers proposed that, rather than transmission of handedness based on the fitness of the genotype, the handedness gene may be located on the sex chromosomes. Their model presumed that left-handedness is genetically recessive, with low penetrance (the allele is present but not expressed) and with a genetic variation located on the X chromosome. Jones and Martin concluded that the chance of being left-handed is recessive rather than additive as assumed by Corvallis and asserted that the distribution of left-handedness is consistent with that which would be expected for a recessive X-linked gene with low phenotypic penetrance.