The Iberian lynx Lynx pardinus is one of the world's most endangered felids and is vulnerable to human-induced mortality and habitat loss, which reduce population size and accelerate the loss of genetic variation. Twenty-five metric traits of Iberian lynx skulls have been measured on 95 skulls collected between 1872 and 2003. The skulls belong to three geographically distinct areas/populations, which have recently diverged from each other as a consequence of increased habitat fragmentation: Doñana area, Sierra Morena mountains and Montes de Toledo area. The morphometric study was undertaken using univariate, multivariate and admixture analysis approaches, and all three techniques provided evidence for morphometric differentiation, both in skull size and shape, among the three populations for both males and females. Environmental and genetic forces that may have shaped these patterns are discussed. The males of the population of the Doñana area showed a different degree of reduction in size in nine of the skull traits with time, which has been suggested to be partly because of worsened habitat conditions. However, the heterogeneity of the degree of mean size reduction and the relatively high degree of reduction of some of the skull traits investigated (>4%), which have altered the original proportions between the skull variables, could also partly be attributed to inbreeding depression in the Doñana population. The phenotypic variability of the skull traits showed significant increases (two traits) or decreases (nine traits) with time, and this different pattern of change with time has been suggested to be because of a different number of genes controlling the traits with different degrees of dominance and epistatic interactions. The increased phenotypic variability of two of the traits has also been attributed to a possible decreased level of developmental stability, which can be produced by environmental and/or genetic stress. The findings of this investigation contribute to the discussion about the utility and the limits of quantitative genetics techniques for conservation purposes.