Morphological integration and modularity have become central concepts in evolutionary biology and geometric morphometrics. This review summarizes the most frequently used methods for characterizing and quantifying integration and modularity in morphometric data: principal component analysis and related issues such as the variance of eigenvalues, partial least squares, comparison of covariation among alternative partitions of landmarks, matrix correlation and ordinations of covariance matrices. Allometry is often acting as an integrating factor. Integration and modularity can be studied at different levels: developmental integration is accessible through analyses of covariation of fluctuating asymmetry, genetic integration can be investigated in different experimental protocols that either focus on effects of individual genes or consider the aggregate effect of the whole genome, and several phylogenetic comparative methods are available for studying evolutionary integration. Morphological integration and modularity have been investigated in many species of mammals. The review gives a survey of geometric morphometric studies in some of the groups for which many studies have been published: mice and other rodents, carnivorans, shrews, humans and other primates. This review demonstrates that geometric morphometrics offers an established methodology for studying a wide range of questions concerning integration and modularity, but also points out opportunities for further innovation.