An examination of the marriage histories in the Canadian Fertility Survey confirm the fact that marriage dissolutions have been increasing rapidly in recent years. Cohort analyses by age and by year of marriage enable 1 to see that this is not just a timing effect and that the probabilities of dissolution in the 1st 10 years of marriage have indeed gone up especially among the young and the more recently married cohorts. Canadian rates are somewhat lower than those found in the US. This study was more concerned with the correlates of marriage dissolution rather than with trends which are better handled with vital statistics. There were very few surprises. Those who marry early were not religious had a premarital birth or lived in a large city had greater chances of separation. Some obviously important correlates such as income or occupation were not included here mainly due to lack of adequate data and given the time-dependent nature of these variables. The findings reveal that when it comes to marriage dissolution the differences according to demographic or socioeconomic characteristics seem to be far more important than say in other demographic events such as childbearing or mortality at least in the present social context in Canada. Conversely it makes little sense to treat the population as homogeneous in analyzing survival probabilities of marriages other than for cursory statistical analviss of global trends. (authors)