Neurotransmitters activate a phospholipase D that is though to specifically hydrolyse phosphatidylcholine. This enzyme has a unique property known as transphosphatidylation: in the presence of an appropriate nucleophilic receptor such as an alcohol, phospholipase D will catalyse the production of phosphatidyl-alcohol. We have studied phospholipase D using an in vitro assay that uses [3H]butanol of high specific radioactivity (15 Ci/mmol) as an acceptor. In the presence of [3H]butanol and phosphatidylcholine, a microsomal membrane fraction from rat brain catalysed the production of phosphatidyl[3H]butanol. Phospholipase D activity was dependent upon the presence of a detergent; the optimal sodium oleate concentration was between 4 and 6 mM. The RF of the phosphatidyl[3H]butanol on t.l.c. was identical to the RF of the phosphatidylbutanol formed when [3H]phosphatidylcholine was incubated with 100 mM butanol. These data confirm the identity of phosphatidyl[3H]butanol. One important advantage of this assay is that the substrate does not need to be labelled. We have used this advantage to examine the substrate specificity of phospholipase D. Microsomal phospholipase D appears to hydrolyse phosphatidylcholine most efficiently. There is a relatively small but significant activity against phosphatidylethanolamine and phosphatidylserine, and there is no significant activity against phosphatidylinositol. As the head-group becomes more like choline, the phospholipid becomes a better substrate for phospholipase D. The addition of one methyl group leads to a large increase in activity. Fatty acid composition does not play a role in determining the substrate specificity. This assay should be useful in furthering our understanding of this important enzyme.