Purpose The present study assesses how sibship size affects child quality as measured by educational attainment. Design/methodology/approach The data are from the Canadian General Social Surveys (GSS) of 1986, 1990, 1994 and 1995. The sample is restricted to the individuals born in Canada between 1946 and 1965, that is, the baby-boom generation. In addition to controlling for parental education, the sibship size is instrumented by a non-binary variable created based on the sex composition of the sibship. While most previous studies have pooled both genders, the present paper produces by gender estimates Findings The OLS estimates are statistically significant, negative and moderately large for both male and female baby boomers. When the sibship size is instrumented, the estimates indicate that one additional sibling had reduced the educational attainment of male baby boomers by almost half a year. No causal effect for the sibship size is found for female baby boomers. Originality/value This is the first paper on the effects of sibship size on educational attainment, using Canadian data.