Serine sulphhydrase from chicken liver and cysteine lyase from chicken-embryo yolk sac catalyse the exchange of alpha-H atoms of the amino acid substrate with 3-H-2O. The degree of labelling of the unreacted substrate approaches a maximum of one atom per mol of amino acid. In the absence of replacing agent there is practically no H-exchange in the substrate. The alpha-H of the accumulating beta-substitution product is completely replaced by the labelled hydrogen of the solvent water, irrespective of the duration of incubation. The amount of labelled alpha-hydrogen incorporated into excess (unreacted) amino acids substrate within 3.5-h incubation is somewhat less than the amount incorporated into the product of the complete enzymic beta-replacement reaction. Within the sensitivity limits of detection, the enzymes do not induce any isotopic exchange either of b-H atoms in the amino substrate or of 18-O-labelled beta-HO groups, in the case of L-serine. Neither serine sulphhydrase nor cysteine lyase will catalyse alpha-hydrogen exchange in close structural analogues of their substrates, e.g. L-alanine, D-serine, threonin, 3-phosphoserine. A special case is the interaction of cysteine lyase with the competitive inhibitor, L-serine (whose inhibitor constant, K-i, is equal to the Michaelis constant, K-m, of L-cysteine): the lyase catalyses, only in presence of a cosubstrate thiol, alpha-H exchange in L-serine at approximately the same rate as in L-cysteine. The present data concerning isotopic alpha-H exchange in substrate amino acids, and evidence published earlier, suggest that the catalytic mechanism of replacement-specific beta-lyases may significantly differ from that of the eliminating or ambivalent (mixed-function) lyases. Formation of alpha, beta-unsaturated pyridoxylidene aldimines as real reaction intermediates is unlikely in the case of lyases specifically catalysing beta-replacement reactions; these may proceed by some alternative mechanism of the type suggested in this paper.