This paper investigates the effect of parents’ current income and long-term family characteristics on individuals’ highest educational qualification obtained by age 26 using UK data from the 1970 British Cohort Study. The issues of the possible sample selection bias produced by the not completely random omission of current family income and that of its potential endogeneity are addressed, using a hot-deck multiple imputation procedure and including an indicator of child ability, respectively. I find evidence that current family income has a statistically significant positive impact on children’s education, although it is one of negligible magnitude. Long-term family characteristics are far more important.