To a first order, the Caribbean plate converges obliquely at ~2 cm/yr toward the North American plate. This transpression is partly accommodated across the island of Hispaniola by the partitioning of motion between a fold-and-thrust belt trending NW-SE, and two E-W left-lateral fault systems located 150 km apart. The southern fault, the Enriquillo-Plantain Garden Fault (EPGF), is morphologically well expressed in western Haiti but its precise geometry in eastern Haiti is debatable. There, Lake Azuei stretches over 20 km in a direction parallel to the fold-and-thrust belt while its southern shoreline strikes EW, parallel to the expected trend of the EPGF. Because of a high sedimentation rate, the history of transpressional deformation should be captured in the lake stratigraphy and, accordingly, we acquired 220 km of multichannel seismic reflection (MCS) profiles across its surface. The survey followed a grid pattern with a spacing of 1.2 km and achieved a penetration of up to 300 m beneath lakebed. Interpretation of the dataset documents two major structures. First, the western side of the lake is occupied by a broad NW-trending monoclinal fold. This fold is cross-cut by a few NW-striking vertical (strike-slip) faults. We propose that this monocline is the surface expression of a SW-dipping blind thrust fault. The progressive steepening of the seismic horizons with depth suggests that it has been continuously active during the deposition of at least 300 m of sediments. The other major structure consists of a ~2 km-wide deformation zone that borders the EW-trending southern shore. This deformation zone is faintly imaged below a shallow gas front. We tentatively propose that it corresponds to a set of fault-propagation folds that are developing ahead of an EW, S-dipping oblique-slip fault. Such a model has been proposed already from three other independent studies involving GPS monitoring, seismological monitoring, and detailed field mapping. It is also supported by CHIRP profiles acquired concurrently with our MCS data and that document folding of the topmost turbidites but a lack of evidence for any stratigraphic offset across faults. Furthermore, a set of en echelon folds in that area are trending EW, while WNW-ESE fold axes would be expected instead above an EW vertical strike-slip fault.