A family handedness study of 2632 families and 8605 offspring was conducted. Of the 2632 parental couples, there were 2123 in which both parents were right-handed (RR), 232 in which the mother was left-handed and the father right-handed (LR), 254 in which the mother was right-handed and the father was left-handed (RL), and 23 in which both parents were left-handed (LL). Results showed some important differences from the composite results of four earlier large scale studies that had also employed the same criterion of handedness (writing hand). These had collectively found that the incidence of left-handed offspring, of both sexes, was significantly lower for RR couples than for LR or RL couples, but not lower than for LL couples. Present results, however, suggest an X-linked pattern of genetic influence on handedness. The LR parents produced significantly more left-handed offspring than did RR couples, and this was particularly the case for sons; but while RL couples produced a higher incidence of left-handed daughters than did RR parents, they failed to produce a higher incidence of left-handed sons than did RR couples. Additionally, the present sample showed a significantly greater incidence of left-handed offspring of LL couples than of RR couples, a finding that, while predicted by genetic-influence theories of handedness, was not found in the composite of previous comparison samples. The finding of sex-linkage is potentially important, but will require replicative studies that take special care to preclude possible biasing factors such as selective volunteering of participants and inaccurate offspring ascriptions of parental (particularly paternal) handedness.