We report a joint experimental and theoretical study on the ultrafast excited state dynamics of allene and a series of its methylated analogues (1,2-butadiene, 1,1-dimethylallene, and tetramethylallene) in order to elucidate the conical intersection mediated dynamics that give rise to ultrafast relaxation to the ground electronic state. We use femtosecond time-resolved photoelectron spectroscopy (TRPES) to probe the coupled electronic-vibrational dynamics following UV excitation at 200 nm (6.2 eV). Ab initio multiple spawning (AIMS) simulations are employed to determine the mechanistic details of two competing dynamical pathways to the ground electronic state. In all molecules, these pathways are found to involve as follows: (i) twisting about the central allenic C–C–C axis followed by pyramidalization at one of the terminal carbon atoms and (ii) bending of allene moiety. Importantly, the AIMS trajectory data were used for ab initio simulations of the TRPES, permitting direct comparison with experiment. For each molecule, the decay of the TRPES signal is characterized by short (30 fs, 52 fs, 23 fs) and long (1.8 ps, 3.5 ps, [306 fs, 18 ps]) time constants for 1,2-butadiene, 1,1-dimethylallene, and tetramethylallene, respectively. However, AIMS simulations show that these time constants are only loosely related to the evolution of electronic character and actually more closely correlate to large amplitude motions on the electronic excited state, modulating the instantaneous vertical ionization potentials. Furthermore, the fully substituted tetramethylallene is observed to undergo qualitatively different dynamics, as displacements involving the relatively massive methyl groups impede direct access to the conical intersections which give rise to the ultrafast relaxation dynamics observed in the other species. These results show that the branching between the “twisting” and “bending” pathways can be modified via the selective methylation of the terminal carbon atoms of allene. The interplay between inertial and potential effects is a key to understanding these dynamical branching pathways. The good agreement between the simulated and measured TRPES confers additional confidence to the dynamical picture presented here.