Patterns of winter irruptions in several owl species apparently follow the 'lack of food' hypothesis, which predicts that individuals leave their breeding grounds in search of food when prey populations do not allow breeding and are too small to ensure survival. Recent analyses, however, suggest an alternative mechanism dubbed the 'breeding success' hypothesis, which predicts that winter irruptions might instead be the result of a very successful breeding season, with a large pool of young birds subsequently migrating south from the breeding grounds. Here we assessed age-class (juvenile vs. non-juvenile) composition of winter irruptive Snowy Owls Bubo scandiacus over a 25-year period (winter 1991-1992 to 2015-2016) between regular (North American Prairies and Great Plains) and irregular wintering areas (northeastern North America) using live-trapped individuals and high-resolution images of individual owls. Our results show that the proportion of juveniles (birds less than 1 year of age) varies considerably annually but is positively correlated with irruption intensity in both regions. In irregular wintering areas, it can constitute the majority (up to more than 90%) of winter irruptive Snowy Owls over a large geographical area. These results are consistent with the idea that large winter irruptions at temperate latitudes are not the result of adults massively leaving the Arctic in search of food after a breeding failure but are more likely to be a consequence of good reproductive conditions in the Arctic that create a large pool of winter migrants.