1. An improved procedure is reported for purification of the amine dehydrogenase from methylamine-grown Pseudomonas AM1 which yielded a product homogeneous by sedimentation and disc-electrophoretic analysis, with molecular weight of 133000. 2. The purified enzyme had absorption maxima at 280 and 430nm. On aging, a third peak appeared at 325nm, and the 430nm peak decreased in intensity. This spectrum was independent of pH. 3. Addition of 2.5mm-semicarbazide, phenylhydrazine, hydrazine or hydroxylamine produced modified spectra with maxima respectively at 400, 440, 395 and 425nm. 4. Aerobic addition of methylamine resulted in a bleaching of the 430nm peak and the appearance of a new one at 325nm. This spectral change was retained after removal of the methylamine by dialysis. The original spectrum could be restored on addition of phenazine methosulphate. 5. Addition of borohydride partially inactivated the enzyme and produced spectral changes similar to those observed with methylamine. Pre-treatment with methylamine prevented the inactivation by borohydride. The degree of inactivation could be increased by alternate phenazine methosulphate and borohydride treatments. 6. The addition of methylamine or borohydride each caused shifts in the fluorescence emission maximum from 348 to 380nm. 7. Lineweaver-Burk plots of reciprocal activity against reciprocal concentration of either of the substrates n-butylamine or phenazine methosulphate were consistent with a mechanism that involves interconversion of two free forms of the enzyme by the two substrates. 8. The enzyme, although spectrally modified, was not inactivated by dialysis against diethyldithiocarbamate, and contained about 0.27 g-atom of copper/mol, with small traces of cobalt, iron and zinc. 9. Conventional methods of resolution did not release the prosthetic group. Heat denaturation after treatment of the enzyme with methylamine liberated a yellow chromophore which did not reactivate resolved aspartate aminotransferase, and whose spectral, electrophoretic and fluorescence properties did not agree with any recognizable pyridoxal derivatives. 10. Despite the inconclusive results with the isolated chromophore, the observations on the enzyme suggest that it may contain a pyridoxal derivative bound as a Schiff's base which is converted into the pyridoxamine form on aerobic treatment with methylamine and reconverted into the pyridoxal form with phenazine methosulphate. 11. The copper detected is probably not involved in the enzyme mechanism, since most copper-chelating agents are not inhibitory, and since the enzyme does not react with oxygen.