Papers in this volume and elsewhere consistently find a strong relationship between children's cognitive abilities and their parents' socio-economic position (SEP). Most studies seeking to explain the paths through which SEP affects cognitive skills suffer from a potentially serious omitted variables problem, as they are unable to account for an important determinant of children's cognitive abilities, namely parental cognitive ability. A range of econometric strategies have been employed to overcome this issue, but in this paper, we adopt the very simple (but rarely available) route of using data that includes a range of typically unobserved characteristics, such as parental cognitive ability and social skills. In line with previous work on the intergenerational transmission of cognitive skills, we find that parental cognitive ability is a significant predictor of children's cognitive ability; moreover, it explains one sixth of the socio-economic gap in those skills, even after controlling for a rich set of demographic, attitudinal and behavioural factors. Despite the importance of parental cognitive ability in explaining children's cognitive ability, however, the addition of such typically unobserved characteristics does not alter our impression of the relative importance of other factors in explaining the socio-economic gap in cognitive skills. This is reassuring for studies that are unable to control for parental cognitive ability.