Inferences about the evolution of continuous traits based on reconstruction of ancestral states has often been considered more error-prone than analysis of independent contrasts. Here we show that both methods in fact yield identical estimators for the correlation coefficient and regression gradient of correlated traits, indicating that reconstructed ancestral states are a valid source of information about correlated evolution. We show that the independent contrast associated with a pair of sibling nodes on a phylogenetic tree can be expressed in terms of the maximum likelihood ancestral state function at those nodes and their common parent. This expression gives rise to novel formulae for independent contrasts for any model of evolution admitting of a local likelihood function. We thus derive new formulae for independent contrasts applicable to traits evolving under directional drift, and use simulated data to show that these directional contrasts provide better estimates of evolutionary model parameters than standard independent contrasts, when traits in fact evolve with a directional tendency. The primary reason to select one class of method over another is thus not that they measure different things but that their estimators exhibit different statistical properties that may be more or less desirable [2]. In this sense, independent contrasts and phylogenetic generalized least squares models are generally favoured over ancestral state reconstruction. Pagel [2] argues that independent contrasts are best suited to the problem of identifying evolutionary correlation coefficients, since directional methods based on a tree with n tips count evolutionary changes on 2(n − 1) internal branches, meaning that "half of the variation that a directional method calculates is redundant because it overlaps with variation already calculated" yielding "results that seem more stable than they actually are", whereas independent contrasts, 3 based on values calculated at n − 1 internal nodes, "make use of all the variance in the data, but in a way that does not count any of it twice". Ackerly It is shown here that independent contrasts and maximum likelihood ancestral state reconstruction not only estimate the same underlying Brownian rate parameter for a univariate trait, but also -in studies of correlated evolution -yield numerically identical regression estimators for the gradient and 4 correlation coefficient of bivariate traits. As a consequence, inferences about correlated evolution derived from maximum likelihood ancestral state estimation are as valid as, and indeed identical to, those derived from independent contrasts procedures. We show that the independent contrast associated with a pair of sibling nodes in a phylogenetic tree can be expressed in terms of the Gaussian local likelihood function of the node that is the direct common ancestor of the pair. It thus transpires that the numerical calculations carried out in generating independent contrasts are identical to those carried out in maximum likelihood ances...