Habitat fragmentation is one of the principal threats to primates. Studies of primates in fragments usually conclude that fragmentation negatively affects some aspect of their biology or ecology. Nevertheless, the definition and quantification of fragmentation vary considerably among studies, resulting in contradictions and results that are difficult to interpret. We here 1) discuss the problems associated with the definition of habitat fragmentation and the ways of measuring it, 2) emphasize the importance of the concepts and methods from landscape ecology and metapopulation theory for the study of primates in fragmented landscapes, and 3) offer recommendations for more precise use of concepts associated with habitat fragmentation from the primates' perspective. When specific knowledge of the study species/population is available, we suggest that the definition of the variables to be measured should be functional from the primates' perspective, based, e.g., on their habitat requirements and dispersal capacity. The distance to the nearest fragment may not be the best way to measure the isolation between populations. Fragmentation per se is a landscape scale process and, hence, landscape scale studies are required to understand how species are distributed across heterogeneous landscapes. Finally, it is important to consider that what happens at the fragment scale could be the consequence of processes that interact at various spatial and temporal scales.