This lecture is aimed at reviewing recent results on the study of fission dynamics. It is well established that fission is a slow process dominated by nuclear viscosity. This is demonstrated by many experimental observations, mainly based on measurements of fission fragment Total Kinetic Energy (TKE), pre-scission light particle, GDR γ-ray multiplicities and energy spectra. Fission time-scale as well as the nature of energy dissipation during the process and its dependence on the nuclear shape and the temperature are the main aspects which have been addressed. In spite of the extensive work, there are still many open questions on the dynamics of the process, mainly due to the lack of constraints on the models and to the different probes used. Intermediate fissily systems are particularly suited for this study as they present comparable cross sections in the fusion-fission and fusion-evaporation channels, allowing to measure observables in both channels, and therefore to further constrain the models. Furthermore, for these systems the path from equilibrium to saddle configuration is expected to dominate with respect to saddle-to-scission one, reducing the complexity of the physical process to be studied. As a case study, we present the study performed on the nucleus ¹³²Ce, which shows the limits of the Statistical Model in accounting for the whole set of observables. We further present the analysis of the data with a 3D Langevin dynamical approach which proves to be capable of reproducing the values of a large set of observables. Our analysis reinforces the fact that extended data sets are essential to achieve reliable simulations to address the open questions on fission dynamics and strongly suggests the use of setups of high efficiency.