Borane-amine adducts have received considerable attention, both as vectors for chemical hydrogen storage and as precursors for the synthesis of inorganic materials. Transition metal-catalyzed ammonia-borane (H3N-BH3, AB) dehydrocoupling offers, in principle, the possibility of large gravimetric hydrogen release at high rates and the formation of B-N polymers with well-defined microstructure. Several different homogeneous catalysts were reported in the literature. The current mechanistic picture implies that the release of aminoborane (e.g., Ni carbenes and Shvo's catalyst) results in formation of borazine and 2 equiv of H2, while 1 equiv of H2 and polyaminoborane are obtained with catalysts that also couple the dehydroproducts (e.g., Ir and Rh diphosphine and pincer catalysts). However, in comparison with the rapidly growing number of catalysts, the amount of experimental studies that deal with mechanistic details is still limited. Here, we present a comprehensive experimental and theoretical study about the mechanism of AB dehydrocoupling to polyaminoborane with ruthenium amine/amido catalysts, which exhibit particularly high activity. On the basis of kinetics, trapping experiments, polymer characterization by (11)B MQMAS solid-state NMR, spectroscopic experiments with model substrates, and density functional theory (DFT) calculations, we propose for the amine catalyst [Ru(H)2PMe3{HN(CH2CH2PtBu2)2}] two mechanistically connected catalytic cycles that account for both metal-mediated substrate dehydrogenation to aminoborane and catalyzed polymer enchainment by formal aminoborane insertion into a H-NH2BH3 bond. Kinetic results and polymer characterization also indicate that amido catalyst [Ru(H)PMe3{N(CH2CH2PtBu2)2}] does not undergo the same mechanism as was previously proposed in a theoretical study.