The generally accepted mode of action of Ni- or Co-Mo/Al2O3 catalysts during hydrocracking of residua is that the catalyst participates in the hydrogenation of aromatics which results in heteroatom removal and CCR conversion. Earlier studies have shown that under hydrocracking conditions with residua, the catalyst loses the ability to hydrogenate aromatics within hours, and thereafter catalysis proceeds by a different mechanism. The established techniques of stable isotope analyses have been used to provide additional support for the proposed mechanism. The isotope studies showed that in the absence of a catalyst, there was no incorporation of gaseous hydrogen into the residuum fraction and only a small amount into the distillates. With a fresh catalyst, hydrogen was incorporated into all fractions initially but hydrogenation of residuum lasted for only 4 h. When bitumen residuum was hydrogenated under mild conditions, the added hydrogen was not lost under hydrocracking conditions, Addition of gaseous hydrogen to the residuum was dependent on the concentration of catalyst for dispersed catalysts, and catalysts previously used under mild conditions in the pilot plant did not show any hydrogen transfer from the gas phase to the residuum fraction. In all cases, carbon-to-carbon bond breaking was correlated with the incorporation of gaseous hydrogen.