The northeastern Caribbean provides a natural laboratory to investigate strain partitioning, its causes and its consequences on the stress regime and tectonic evolution of a subduction plate boundary. Here, we use GPS and earthquake slip vector data to produce a present-day kinematic model that accounts for secular block rotation and elastic strain accumulation, with variable interplate coupling, on active faults. We confirm that the oblique convergence between Caribbean and North America in Hispaniola is partitioned between plate boundary parallel motion on the Septentrional and Enriquillo faults in the overriding plate and plate-boundary normal motion at the plate interface on the Northern Hispaniola Fault. To the east, the Caribbean/North America plate motion is accommodated by oblique slip on the faults bounding the Puerto Rico block to the north (Puerto Rico subduction) and to the south (Muertos thrust), with no evidence for partitioning. The spatial correlation between interplate coupling, strain partitioning and the subduction of buoyant oceanic asperities suggests that the latter enhance the transfer of interplate shear stresses to the overriding plate, facilitating strike-slip faulting in the overriding plate. The model slip rate deficit, together with the dates of large historical earthquakes, indicates the potential for a large (M w 7.5 or greater) earthquake on the Septentrional fault in the Dominican Republic. Similarly, the Enriquillo fault in Haiti is currently capable of a M w 7.2 earthquake if the entire elastic strain accumulated since the last major earthquake was released in a single event today. The model results show that the Puerto Rico/Lesser Antilles subduction thrust is only partially coupled, meaning that the plate interface is accumulating elastic strain at rates slower than the total plate motion. This does not preclude the existence of isolated locked patches accumulating elastic strain to be released in future earthquakes, but whose location and geometry are not resolvable with the present data distribution. Slip deficit on faults from this study are used in a companion paper to calculate interseismic stress loading and, together with stress changes due to historical earthquakes, derive the recent stress evolution in the NE Caribbean.