The water entry of a rectangular plate with a high horizontal velocity component is investigated experimentally. The test conditions are representative of those encountered by aircraft during emergency landing on water and are given in terms of three main parameters: horizontal velocity, approach angle, i.e. vertical to horizontal velocity ratio, and pitch angle. Experimental data are presented in terms of pressure, spray root shape, pressure peak propagation velocity and total loads acting on the plate. A theoretical solution of the plate entry problem based on two-dimensional and potential flow assumptions is derived and is used to support the interpretation of the experimental measurements. The results indicate that, as the plate penetrates and the ratio between the plate breadth and the wetted length measured on the longitudinal plane diminishes, the role of the third dimension becomes dominant. The increased possibility for the liquid to escape from the lateral sides yields a reduction of the pressure peak propagation velocity and, consequently, of the corresponding pressure peak intensity. In particular, it is shown that, at the beginning of the entry process, the pressure peak moves much faster than the geometric intersection between the body and the free surface, but at a later stage the two points move along the body at the same speed. Furthermore, it is shown that the spray root develops a curved shape which is almost independent of the specific test conditions, even though the initial growth rate of the curvature is higher for larger pitch angles. The loads follow a linear increase versus time, as predicted by the theoretical model, only in a short initial stage. Next, for all test conditions examined here, they approach a square-root dependence on time. It is seen that, if the loads are scaled by the square of the velocity component normal to the plate, the data are almost independent of the test conditions.