Lynx canadensis (Canada Lynx) mostly occurs in the continental area of North America. Two populations in Atlantic Canada on Newfoundland and Cape Breton Island are geographically isolated. Past studies have revealed geographical and environmental barriers that have significantly impacted processes that ultimately influence the ecology, genetics, evolution, and conservation of the species' populations. However, equivocal results were obtained as to the morphological and genetic characteristics of the species, and very little is known in this regard on the island populations. The aim of this study was to investigate skull morphometric variation between the species' populations. We examined and measured 18 craniodental characters on 171 specimens spanning the species' Canadian range, including most of its boreal forest range and the 2 island populations. Univariate and multivariate analyses provided evidence for significant morphological differentiation among the species' populations. Factors pertaining to geographical isolation of populations accounted for the bulk of the craniometric variation for both males and females. Principal component analysis (PCA) identified 3 geographical groups: mainland Canada, Newfoundland, and Cape Breton Island. Consistent with the “island rule”, analysis of variance with the Sheffé significant difference post hoc test and PCA results indicated that continental individuals were significantly larger than those from Cape Breton Island, whereas those from Newfoundland, which is significantly larger than Cape Breton Island, exhibited intermediate size. Shape-related variations in the frontal bone among the 3 geographical groups were detected. Six out of the 18 craniodental characters, i.e., postorbital constriction, mastoid, mandible, upper tooth row, mandibular molar row, and upper canine to canine segment, can be used in discriminant function analysis to distinguish between these 3 groups and correctly classify more than 90% of the individuals from the insular populations. We showed for the first time differentiation between Canada Lynx populations on Cape Breton Island and those in the rest of Canada. Individuals from Cape Breton Island appeared the smallest. Contrary to expectations, the Rocky Mountains did not prove to be a significant geographical barrier, resulting in no morphological differentiation of the British Columbia populations. The morphological variations we reported in this study should benefit conservation and management programs in Atlantic Canada, where the levels of the Canada Lynx population are critically low. Conservation plans that strive to maintain the genetic variation documented in the present study should help ensure the preservation of sufficient variability for adaptation to changing environmental conditions required for the long-term viability of populations of Canada Lynx.