This research examines factors associated with the timing of first birth in Canada, focusing primarily on the role of women's educational attainment. Proportional hazards modelling techniques are applied to data from the 1984 Canadian Fertility Survey (CFS) in order to determine how educational attainment, estimated as close as possible to the date of first birth, influences the timing of first birth and whether the importance of this variable varies according to age cohorts. The results suggest that among a number of variables useful for distinguishing different levels of risk, educational attainment proves to be the most important predictor in the model. Education exerts a substantial positive influence on birth timing for women of all age groups. As expected, moreover, significant cohort differences are also evident, with the greatest to the smallest impact on the risk from the youngest to the oldest cohorts. These clear-cut cohort differences indicate a fundamental change in the effects of education over time, a trend most likely resulting from substantial changes in both the content and social significance of formal schooling during the past few decades.