Many biodiversity conservation strategies aim to increase species movement by changing the landscape between suitable areas of habitat. We applied systematic review and meta-analytical methods to robustly assess evidence on the impact of matrix structure on movement rates, with the hypothesis that movement will be greater through matrix of a more similar structure (vegetation height and cover) to the home habitat. Twenty studies of movement through two or more different matrix types provided 107 effect sizes, expressing the difference between the relative movement rates in different matrix types. The studies were all on animals, including relatively mobile taxa such as birds and butterflies but also rodents and amphibians. We were able to detect that on average, movement was greater through matrix of a more similar structure to the species' habitat despite the variation in studies in terms of matrix types, species and methods. The effect size was larger when there was a greater difference in the structure of the two matrix types being compared (e.g. comparing grassland to forest rather than short grass to long grass). However, there was a high degree of covariation between matrix contrast and studies and other significant subgroupings such as taxonomic group and matrix openness. The biological significance of the increase in movement is not clear; however, ecological theory predicts dispersing individuals are important for population dynamics. Changes to the structure of landscapes intended to improve permeability to movement are supported by the findings of this study, particularly for relatively mobile species. However, research over longer timescales, greater distances and range of taxonomic groups is necessary.