The fact that biotin synthase, fromEscherichia coliandBacillus sphaericus,requiresS-adenosylmethionine and a reducing system led us to postulate that this synthase could belong to the family of enzymes which useS-adenosylmethionine as a source of deoxyadenosyl radical, namely pyruvate formate-lyase, lysine 2,3-aminomutase, and anaerobic ribonucleotide reductase. We describe here experiments withS-[2,8-3H] adenosylmethionine andS-adenosyl-[methyl-3H]methionine which allowed the identification and quantification of the expected cleavage products, deoxyadenosine, and methionine. They are formed in equimolar amounts, in a ratio close to 3 with respect to the biotin produced. We postulate a mechanism involving the homolytic cleavage of two C–H bonds which should consume two equivalents ofS-adenosylmethionine. The observed excess ofS-adenosylmethionine consumption is attributed to abortive processes.