Coevol is a Bayesian inference program using Markov Chain Monte Carlo methods. It can be seen as a fusion between classical phylogenetic models for nucleotides or codons (Felsenstein, 1981), autocorrelated relaxed clocks for molecular dating (Thorne et al., 1998) and compar- ative models based on the idea of Brownian processes and phylogenetically independent contrasts (Felsenstein, 1985; Martins and Hansen, 1997; Harvey and Pagel, 1991). In Coevol, the estimation works by conditioning a probabilistic model simultaneously on a sequence alignment, a matrix of quantitative characters (such as morphological data or life- history traits) and fossil calibrations. The model assumes correlated evolution of the rate of substitution, or other parameters of the substitution process such as GC content or dN/dS, and the quantitative traits, all of which are jointly modeled as a multivariate Brownian process. The program then estimates the correlation structure between these variables (i.e. the covariance matrix of the Brownian process) and simultaneously reconstructs divergence times and ancestral trait values along the phylogeny.