Quasi-likelihood methods are a flexible way of modeling multivariate data without specifying the full distribution. They only require the specification of the mean and variance functions, and a dispersion parameter. These methods have been introduced by Wedderburn (Biometrika 61:439–447, 1974), but have not received much attention in the context of multivariate models, such as GEE models and GLMM. This chapter will present the extension of quasi-likelihood methods to multivariate data, and discuss the quasi-likelihood functions, estimating equations, quasi deviances, and model tests based on them.