Solubilized rhesus monkey liver microsomes were used as the starting material for the purification of epoxide (cis-stilbene oxide) hydrolase. Successive chromatography over DEAE-Sephacel followed by CM-cellulose resulted in two peaks of activity, CM A and CM B. Passage of these two eluates over separate hydroxyapatite columns resulted in two peaks of activity from CM A, HA A1, and HA A2, and one peak from CM B and HA B, with respective recoveries of 1, 7, and 0.2% of cis-stilbene oxide hydrolase activities. A similar recovery was found for benzo[a]pyrene-4,5-oxide hydrolase, while trans-stilbene oxide hydrolase activity coeluted only in HA A2. Fraction HA A1 was homogeneous as judged by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Immunoblots of the three eluates and solubilized microsomes incubated with anti-HA A1 demonstrated a single band at 49 kDa in each fraction. The three eluates were differentially affected by the inhibitors of epoxide hydrolase, trichloropropene oxide and 4-phenylchalcone oxide, and addition of Lubrol PX and phospholipid. Immunoprecipitation of HA A2 resulted in coprecipitation of cis- and trans-stilbene oxide hydrolase activity. Upon immunoprecipitation of solubilized microsomes, all the cis-stilbene oxide and benzo[a]pyrene-4,5-oxide, but only 50-60% of trans-stilbene oxide hydrolase activity was precipitated. These studies support findings with other species that (i) an immunochemically distinct cytosolic-like epoxide hydrolase exists in microsomes, and (ii) microsomal epoxide hydrolase activity can be separated during ion-exchange chromatography giving proteins with similar molecular weights and immunochemical cross-reactivity. The precipitation of cis- and trans-stilbene oxide hydrolase activity in eluate HA A2 provides convincing evidence that these isozymes are not structurally identical.