Mushroom tyrosinase, which is known to convert a variety of o-diphenols into o-benzoquinones, has been shown to catalyse an unusual oxidative decarboxylation of 3,4-dihydroxymandelic acid to 3,4-dihydroxybenzaldehyde [Sugumaran (1986) Biochemistry 25, 4489-4492]. The mechanism of this reaction was re-investigated. Although visible-region spectral studies of the reaction mixture containing 3,4-dihydroxymandelic acid and tyrosinase failed to generate the spectrum of a quinone product during the steady state of the reaction, both trapping experiments and non-steady-state kinetic experiments provided evidence for the transient formation of unstable 3,4-mandeloquinone in the reaction mixture. The visible-region spectrum of mandeloquinone resembled related quinones and exhibited an absorbance maximum at 394 nm. Since attempts to trap the second intermediate, namely alpha,2-dihydroxy-p-quinone methide, were in vain, mechanistic studies were undertaken to provide evidence for its participation. The decarboxylative quinone methide formation from 3,4-mandeloquinone dictates the retention of a proton on the alpha-carbon atom. Hence, if we replace this proton with deuterium, the resultant 3,4-dihydroxybenzaldehyde should retain the deuterium present in the original substrate. To test this hypothesis, we chemoenzymically synthesized alpha-deuterated 3,4-dihydroxymandelic acid and examined its enzymic oxidation. Our studies reveal that the resultant 3,4-dihydroxybenzaldehyde retained nearly 90% of the deuterium, strongly indicating the transient formation of quinone methide. On the basis of these findings it is concluded that the enzymic oxidation of 3,4-dihydroxymandelic acid generates the conventional quinone product, which, owing to its unstability, is rapidly decarboxylated to generate transient alpha,2-dihydroxy-p-quinone methide. The coupled dienone-phenol re-arrangement and keto-enol tautomerism of this quinone methide produce the observed 3,4-dihydroxybenzaldehyde.