A method, involving paper chromatography and [14C]cocaine, was developed which measures the hydrolysis of [mu]mole quantities of ([long dash])-cocaine and other tropane alkaloids in serum or organ homogenates of different animals. All previous authors have assumed that the enzymatic hydrolysis of the diester, cocaine, commences with the hydrolysis of the methyl alcohol residue, because O-benzoyl-([long dash])-ecgonine is not a substrate of "cocaine esterase". On the basis of our studies, this enzyme should be called "([long dash])-cocaine-3-acylhydrolase". Three samples of cocaine were used, labelled at different positions in the molecule ([N-HCH3J-, [14COC6H5]- and ([long dash])-[CO2l4CH4]cocaine). With these substrates it was shown that in rabbit serum containing "([long dash])-cocaine-3-acylhydrol-ase", the benzoic acid is removed first with the formation of ([long dash])-ecgonine-methyl ester. At the same time a relatively insignificant amount of the methyl alcohol group is hydrolysed non-enzymatically, with the formation of O-benzoyl-ecgonine and ecgonine. The hydrolysis of [14c]cocaine is quite different in rabbit serum containing no "(-[long dash])-cocaine-3-acyl-hydrolase". Under our test conditions the methyl alcohol group is hydrolysed, giving chiefly O-benzoyl-([long dash])-ecgonine. This is then converted very slowly to ecgonine by removal of the benzoic acid (35% in 20 days). Sixty four rabbits were investigated and 4 had defective "([long dash])-cocaine-3-acylhydrolase". The sera of 12 other animal species, and human serum and liver homogenate showed no enzymatic activity towards ([long dash])-cocaine. The activity of "([long dash])-cocaine-3-acylhydrolase" in the individual sera of 16 rabbits varied between 0.064 and 0.188 enzyme units (U)/ml. In rabbit serum, under our experimental conditions, the benzoic acid produced from the ([long dash])-cocaine by hydrolysis was not converted into hippuric acid.