Doubly reduced 9,10-dihydro-9,10-diboraanthra-cenes (DBAs) are introduced as catalysts for hydrogenation as well as hydride-transfer reactions. The required alkali metal salts M-2[DBA] are readily accessible from the respective neutral DBAs and Li metal, Na metal, or KC8. In the first step, the ambiphilic M-2[DBA] activate H-2 in a concerted, metal-like fashion. The rates of H-2 activation strongly depend on the B-bonded substituents and the counter cations. Smaller substituents (e.g., H, Me) are superior to bulkier groups (e.g., Et, pTol), and a Mes substituent is even prohibitively large. Li+ ions, which form persistent contact ion pairs with [DBA](2-), slow the H-2-addition rate to a higher extent than more weakly coordinating Na+/K+ ions. For the hydrogenation of unsaturated compounds, we identified Li-2[4] (Me substituents at boron) as the best performing catalyst; its substrate scope encompasses Ph(H)C=NtBu, Ph2C=CH2, and anthracene. The conversion of E-Cl to E-H bonds (E = C, Si, Ge, P) was best achieved by using Na-2[4]. The latter protocol provides facile access also to Me2Si(H)Cl, a most important silicone building block. Whereas the H-2-transfer reaction regenerates the dianion [4](2-) and is thus immediately catalytic, the H--transfer process releases the neutral 4, which has to be recharged by Na metal before it can enter the cycle again. To avoid Wurtz-type coupling of the substrate, the reduction of 4 must be performed in the absence of the element halide, which demands an alternating process management (similar to the industrial anthraquinone process).