The large 2400 MWth Gas-cooled Fast Reactor (GFR) is one of the most promising advanced reactor concepts currently being investigated within Generation IV. The current work presents the three-dimensional simulation and analysis of GFR core behavior related, primarily, to control assembly (CA) fast movements and/or to accidental withdrawals/ejections. These two categories of events differ in the speed at which the CA is removed. The transient analysis is focused on two different GFR core designs: (1) the reference design ("2004-Core") and (2) a variant design ("2007-Core"), the main difference being in the core geometry, i.e., in the height-to-diameter (H/D) ratio (2.35 m for the latter core, compared to 1.55 m for the reference design). The reported work firstly concerns the development and benchmarking of coupled, full-core 3D neutron kinetics (NK) and 1D thermal-hydraulics (TH) models for both GFR designs. The models were developed using the TRACE and PARCS codes, which form part of the FAST code system applied at PSI for the generic analysis of advanced fast-spectrum systems. The neutronics is simulated in the nodal diffusion approximation, in conjunction with spatial kinetics. In the TH model, each of the fuel subassemblies (SAs) is modeled individually, so that one can accurately track changes in the thermal-hydraulic-channels surrounding the control assemblies. Consistent nodalization schemes are used in the TH and NK models, in order to provide a 1:1 mapping. Using the developed coupled models, spatial effects have been studied for a variety of hypothetical CA driven transients. Asymmetric CA ejections (speed ∼ 4 cm/s) and withdrawals (speed ∼ 2 mm/s) were analyzed, and large spatial deformations of the power map (up to +21%, relative to average fission power in the TH channels adjacent to the ejected CA) were observed. The analysis has clearly brought out the importance of 3D spatial effects, in the context of the correlation between CA-worths and the loose neutronic coupling of core regions in the case, for example, of a low-H/D core. A sensitivity study of the spatial effects to different parameters (e.g., number of assemblies being ejected and the CA speed) has been conducted, in order to fully characterize the GFR dynamic response. Furthermore, comparisons between the GFR core variants were made. Lower temperatures (fuel, clad and coolant) and lower spatial effects were obtained for the second core. Based upon the results obtained, no major safety-related problem has been found as regard to the CA pattern developed for the GFR. In order to limit the spatial effects, the CAs have to be moved either independently or in symmetric configurations, e.g., three CAs with 120°C symmetry.