Background A history of past major depressive episodes (MDEs) is known to be a risk factor for future MDEs. Additional information about the relationship between past and future episodes would be useful in clinical practice, it is therefore important to fully understand the epidemiology of major depression. We asked whether the number of previous MDEs is related to the probability of recurrence in the general population. Methods Data were used from the Canadian National Population Health Survey (NPHS) that was repeated every 2 years from 1994/1995 to 2009/2010 (i.e., nine cycles). Prior year depression was assessed with the Composite International Diagnostic Interview Short Form (CIDI‐SF). We estimated the total number of MDEs in individuals over the first eight cycles and examined recurrence in the ninth cycle. These analyses employed a generalized linear model (identity link) where recurrence in cycle 9 was the outcome and the predictor variables were age, gender, and the number of MDEs in the first eight cycles. Results The risk for recurrence of depression in cycle 9 was found to progressively increase with the number of prior episodes, reaching a value of greater than 46% when the number of prior episodes was five to eight. Independent of this association, the risk of recurrence was greater in younger people and women, but the strength of association of these variables was much weaker for past episodes. Conclusions MDE recurrence strongly depended on the number of preceding episodes. Those at highest risk of recurrence can be easily identified by their number of past episodes.