Mammalian soluble and microsomal epoxide hydrolases have been proposed to belong to the family of alpha/beta-hydrolase-fold enzymes. These enzymes hydrolyse their substrates by a catalytic triad, with the first step of the enzymatic reaction being the formation of a covalent enzyme-substrate ester. In the present paper, we describe the direct visualization of the ester formation between rat microsomal epoxide hydrolase and its substrate. Microsomal epoxide hydrolase was precipitated with acetone after brief incubation with [1-(14)C]epoxystearic acid. After denaturing SDS gel electrophoresis the protein-bound radioactivity was detected by fluorography. Pure epoxide hydrolase and crude microsomes showed a single radioactive signal of the expected molecular mass that could be suppressed by inclusion of the competitive inhibitor 1,1,1-trichloropropene oxide in the incubation mixture. In a similar manner, 4-fluorochalcone-oxide-sensitive binding of epoxystearic acid to rat soluble epoxide hydrolase could be demonstrated in rat liver cytosol. Under similar conditions, no covalent binding of [26-(14)C]cholesterol-5alpha,6alpha-epoxide to microsomal proteins or solubilized fractions tenfold enriched in cholesterol epoxide hydrolase activity could be observed. Our data provide definitive proof for the formation of an enzyme-substrate-ester intermediate formed in the course of epoxide hydrolysis by microsomal epoxide hydrolase, show no formation of a covalent intermediate between cholesterol epoxide hydrolase and its substrate under the same conditions as those under which an intermediate was shown for both microsomal and soluble epoxide hydrolases and therefore indicate that the cholesterol epoxide hydrolase apparently does not act by a similar mechanism and is probably not structurally related to microsomal and soluble epoxide hydrolases.