<jats:title>Abstract</jats:title> <jats:p> Experiments with cerulenin-inhibited cultures of 5. violaceoruber showed conversion of dihydrogranaticin (II) into granaticin (I), but not vice versa, confirming an earlier conclusion that II is the biosynthetic precursor of I. Feeding of CH<jats:sub>3</jats:sub> <jats:sup>13</jats:sup>C<jats:sup>18</jats:sup>O<jats:sub>2</jats:sub>Na followed by <jats:sup>13</jats:sup>C-NMR analysis of the product by the <jats:sup>18</jats:sup>O shift method indicated the expected incorporation of <jats:sup>18</jats:sup>O at carbons 1,11 and 13 of I and showed that the oxygen of the pyran ring originates from C-3 and not from C-15. Analysis of I biosynthesized from <jats:sup>13</jats:sup>C<jats:sup>2</jats:sup>H<jats:sub>3</jats:sub>COONa by <jats:sup>13</jats:sup>C{<jats:sup>1</jats:sup>H, <jats:sup>2</jats:sup>H} triple resonance NMR spectroscopy showed the incorporation of one atom of deuterium each at C-2 and C-4. C-16 carried a maximum of 2, not 3, atoms of deuterium. These results are discussed in terms of biosynthetic mechanisms.