This essay advances the definition of Caribbean studies as a practice. Rather than understanding this field of study as a preconceived category open for arbitrary self-appointment, a Caribbeanist practice should be something "which in fact happens (and can be shown to have happened)"—to refer to E. P. Thompson's dictum. It "happens" through the multidisciplinary and border-crossing research of the Caribbean establishing intra- and extraregional connections that enable the better understanding of Caribbean processes. The article begins by reviewing the ideas on the field advanced by several pioneering scholars. It then provides two concrete examples of the author's own research practice: one of a violent incident in the Cuban Republican era and another of intra-Caribbean migration processes in the late nineteenth and early twentieth centuries. As a conclusion, the author reconsiders the relation of Caribbean studies with other geographical frameworks of analysis and fields of study.