This is a review of various aspects of estimation of genetic parameters, concentrating on developments since the last World Congress. Improvements in algorithms to estimate variance and covariance components are described. We discuss the relevance of these procedures, developed for unselected populations, when they are applied to selected populations and indicate unresolved d iffic u ltie s . We comment on the use of mixed model methodology to estimate realised h e rita b ilities. We consider the design of experiments to estimate multivariate parameters. In particular, we. show that for bivariate uncorrelated data the variance of the estimated genetic parameters can be reduced by 0.4 relative to more conventional designs.