<jats:title>Abstract</jats:title><jats:p>A polyene macrolide antibiotic tetramycin biosynthetic gene cluster was identified by genome mining and isolated from <jats:italic>Streptomyces hygrospinosus</jats:italic> var. <jats:italic>beijingensis</jats:italic>. Genetic and in silico analyses gave insights into the mechanism of biosynthesis of tetramycin, and a model of the tetramycin biosynthetic pathway is proposed. Inactivation of a cytochrome P450 monooxygenase gene, <jats:italic>tetrK</jats:italic>, resulted in the production of a tetramycin B precursor: tetramycin A, which lacks a hydroxy group in its polyol region. TetrK was subsequently overexpressed heterologously in <jats:italic>E. coli</jats:italic> with a His<jats:sub>6</jats:sub> tag, and purified TetrK efficiently hydroxylated tetramycin A to afford tetramycin B. Kinetic studies revealed no inhibition of TetrK by substrate or product. Surprisingly, sequence‐alignment analysis showed that TetrK, as a hydroxylase, has much higher homology with epoxidase PimD than with hydroxylases NysL and AmphL. The 3D structure of TetrK was then constructed by homology modeling with PimD as reference. Although TetrK and PimD catalyzed different chemical reactions, homology modeling indicated that they might share the same catalytic sites, despite also possessing some different sites correlated with substrate binding and substrate specificity. These findings offer good prospects for the production of improved antifungal polyene analogues.