[en] The Surrogate nuclear reactions method, an indirect approach for determining cross sections for compound-nuclear reactions involving difficult-to-measure targets, is reviewed. Focusing on cross sections for neutron-induced reactions on actinides, we review the successes of past and present applications of the method and assess its uncertainties and limitations. The approximations used in the analyses of most experiments work reasonably well for (n,f) cross sections for neutron energies above 1-2 MeV, but lead to discrepancies for low-energy (n,f) reactions, as well as for (n,γ) applications. Correcting for some of the effects neglected in the approximate analyses leads to improved (n,f) results. We outline steps that will further improve the accuracy and reliability of the Surrogate method and extend its applicability to reactions that cannot be approached with the present implementation of the method.