Rat brain ceramidase, the enzyme that catalyzes both the hydrolysis and synthesis of ceramide (A-acy lsphingosine), was purified over 200-fold. Advantage was taken of the fact that the enzyme withstands prolonged treatment with trypsin and chymotrypsin. This treatment digests 80% of the protein and decreases the molecular weight of the enzyme as determined by gel filtration through Bio-Gel but does not impair the enzymatic activity. Evidence is presented that fatty acyl coenzyme A is not a direct substrate for ceramide synthesis. The acyl portion can be utilized only after hydrolysis by an accompanying hydrolase to coenzyme A and a free fatty acid. Trials were made to separate the hydrolytic and synthetic activities by subjecting the enzyme to ammonium sulfate fractionation, sonic irradiation, adjustment to acid pH, heating, treatment with proteolytic enzymes or with SH reagents, and chromatography on TEAE-cellulose. None of these procedures resulted in a separation of the two activities from each other. Reaction mixtures were taken to an apparent equilibrium, starting with either ceramide or a mixture of sphingosine and fatty acid. The calculated equilibrium constant, defined as Kequil = (sphingosine) × (fatty acid)/(ceramide), depended upon the substrate employed. It was about 10-4 m when determined in the direction of ceramide synthesis, but only 5 × 10-6 m when measured in the direction of hydrolysis. The meaning and possible significance of these findings are discussed.