Radioactive mevalonic acid, in which the radioactivity was due solely to 5S tritiated molecules, was injected intraperitoneally into living rats. After 45 min the animals were killed and radioactive cholesterol was isolated from the livers. This cholesterol was incubated with Myco-bacterium phlei in conditions that forced accumulation of androst-1,4-dien-3,17-dione. After removal of alkali-labile tritium from the 16$\beta $ position the dione (A) was incubated with Aspergillus tamarii to yield androst-1,4-dien-11$\beta $-ol-3,17-dione, oxidized by chromic acid to androst-1,4-dien-3,11,17-trione (B). When this was treated with alkali the recovered trione (C) was almost non-radioactive. The ratio of the specific radioactivities A:B:C was 1.00:0.51:0.004. It was concluded that within experimental error the cholesterol was equally labelled with tritium at the 11$\beta $ and 12$\alpha $ positions and therefore that both double bonds in the isotopically asymmetric intermediate squalene were epoxidized to an equal extent. It follows that squalene, when biosynthesized by intact rat liver in vivo, is not transferred in a spatially oriented manner from the enzyme that produces it to the enzyme that epoxidizes it.