It has been suggested that the long-standing association of variability in the human nasal index [100 × (nasal breadth)/(nasal height)] with climatic variation is spurious evidence for natural selection in humans (Hoyme, 1965; St. Hoyme and Iscan, 1989). The argument is based principally on the observation that nasal height is globally more variable than nasal breadth, with nasal breadth thus contributing little to variation in the index. This argument does not take into account the confounding effect of absolute size of these variables on their variances. In this study we compare the intrinsic variation in skeletal nasal height and breadth within and among 26 mixed-sex populations (N = 2,408) at globally diverse localities (Howells, 1989), using 2 × 2 variance-covariance matrices of the logarithmically transformed variates. Hypothesis tests for homogeneity of matrices and equalvariance/equal-covariance indicate that the intrinsic variation in nasal breadth is greater than that for nasal height within populations, and that nasal breadth and nasal height exhibit equivalent intrinsic variation among populations. The argument that nasal breadth contributes little to the worldwide variation in the human nasal index is rejected. Given our present understanding of nasal physiological morpho-function, these results support, but do not demonstrate, an adaptive role for human nasal index variation. Promising methods for elucidating natural selection on human nasal form are suggested.