The mechanism of the phosphoglucomutase from Micrococcus lysodeikticus was investigated. Induced-transport tests at low substrate concentrations (0.15mm) showed co-transport of the (32)P label but no induced transport of the (14)C label, which is in quantitative agreement with a phosphoenzyme mechanism with a rapid isomerization of the phosphoenzyme. The results excluded an intramolecular transfer of phosphate and could only have been compatible with a sequential mechanism if the K(m) for glucose 1-phosphate had been over 20 times smaller than the measured value. The results of induced-transport tests at intermediate concentrations (1mm) with both labels agreed quantitatively with a phosphoenzyme mechanism, and induced-transport tests with (14)C-labelled substrates at high concentrations (26mm) indicated that the rate constants for isomerization of the phosphoenzyme must be greater than about 3x10(6)s(-1). Consistent with these findings is the fact that (14)C label exchanged between the substrates twice as rapidly as the (32)P label at chemical equilibrium. Further, since the (14)C label exchanged between the substrates about ten times more rapidly than between the substrates and glucose 1,6-diphosphate, glucose 1,6-diphosphate is not an obligatory intermediate in the interconversion of the substrates. It is concluded that, contrary to previous evidence, the mechanism of the enzyme from M. lysodeikticus is essentially that of the rabbit muscle enzyme. To account for the rapid isomerization of the phosphoenzyme in both cases a mechanism is proposed in which there is no formal isomerization of the phosphoenzyme.