Three routes were explored for their potential to produce the aromatic aldehydes benzaldehyde, p-hydroxybenzaldehyde, and vanillin, which are important flavor components. The first was to exploit organisms which degrade aromatic amino acids via the cinnamate pathway. Trichosporon beigelii was selected and good evidence collected that benzaldehyde was a degradative intermediate. While accumulation of benzoic acid from phenylalanine or cinnamic acid was demonstrated, the aldehyde was not accumulated under any of the conditions examined due to the very high activity of benzaldehyde dehydrogenase. An analogous situation appeared to exist with ferulic acid, from which only vanillic acid was accumulated. Using fungi of the white rot type, reduction of benzoic acid and vanillic acid to benzaldehyde and vanillin was achieved, but rates were low and yields poor due to further reduction to the alcohol. The third and most successful route was microbial conversion of aromatic amino acids to the corresponding phenylpyruvic acids, which were then readily converted to the aldehyde with mild base. A range of organisms were identified which, under appropriate conditions, were effective producers of phenylpyruvic acids. Proteus vulgaris in particular, transformed phenylalanine, tyrosine, or methoxytyrosine into the corresponding phenylpyruvic acids, which could then be converted to benzaldehyde, p-hydroxy-benzaldehyde, and vanillin respectively. The overall process as optimized for benzaldehyde was rapid, relatively simple, and high-yielding.