The potential of two Rhodococcus strains for biotechnological vanillin production from ferulic acid and eugenol was investigated. Genome sequence data of Rhodococcus sp. I24 suggested a coenzyme A-dependent, non-beta-oxidative pathway for ferulic acid bioconversion, which involves feruloyl-CoA synthetase (Fcs), enoyl-CoA hydratase/aldolase (Ech), and vanillin dehydrogenase (Vdh). This pathway was proven for Rhodococcus opacus PD630 by physiological characterization of knockout mutants. However, expression and functional characterization of corresponding structural genes from I24 suggested that degradation of ferulic acid in this strain proceeds via a beta-oxidative pathway. The vanillin precursor eugenol facilitated growth of I24 but not of PD630. Coniferyl aldehyde was an intermediate of eugenol degradation by I24. Since the genome sequence of I24 is devoid of eugenol hydroxylase homologous genes (ehyAB), eugenol bioconversion is most probably initiated by a new step in this bacterium. To establish eugenol bioconversion in PD630, the vanillyl alcohol oxidase gene (vaoA) from Penicillium simplicissimum CBS 170.90 was expressed in PD630 together with coniferyl alcohol dehydrogenase (calA) and coniferyl aldehyde dehydrogenase (calB) genes from Pseudomonas sp. HR199. The recombinant strain converted eugenol to ferulic acid. The obtained data suggest that genetically engineered strains of I24 and PD630 are suitable candidates for vanillin production from eugenol.