The relations between the origins of public educational systems and school attendance remain far from clear. For instance, the proportion of children receiving some sort of formal education did not increase automatically with the extension and elaboration of school facilities. Carl Kaestle has argued that proportionally as many children attended school in New York City in 1750 as in 1850. Elsewhere he and Maris Vinovskis have shown the surprisingly high rate of attendance of children in rural New York and Massachusetts in the early nineteenth century, prior to the so-called common school revival. And our own work has demonstrated that in at least one industrializing city the upward curve of school attendance among adolescent young people was not secular. School attendance, of course, was a differential process. It varied according to place, age, sex, class and ethnicity. However, the exact nature of that variation still has not been delineated. In an earlier article Katz showed why it is important to understand patterns of school attendance and outlined some of the principal ones he had uncovered in Hamilton, Ontario, in 1851 and 1861. Davey developed the analysis much further in his study of working-class school attendance in Hamilton in the latter part of the nineteenth-century. However, both examinations used entirely descriptive statistics, primarily cross-tabulation, and it is possible that some of the relations we uncovered appeared significant only because we had not controlled for other variables. Moreover, our method made it difficult for us to rank with certainty the importance of the various factors that affected school attendance. One major study by Kaestle and Vinovskis has attempted to unravel the complexities of school attendance with multivariate analysis, but their results are rather inconclusive. Indeed, the principal finding is the rather obvious point that age was the most important factor.