Ubiquinone (Coenzyme Q) is a vital respiratory cofactor and antioxidant in eukaryotes. The recent discovery that kaempferol serves as a precursor for ubiquinone's benzenoid moiety both challenges the conventional view of flavonoids as specialized metabolites, and offers new prospects for engineering ubiquinone in plants. Here, we present evidence that Arabidopsis thaliana mutants lacking kaempferol 3-O-rhamnosyltransferase (ugt78d1) and kaempferol 3-O-glucosyltransferase (ugt78d2) activities display increased de novo biosynthesis of ubiquinone and increased ubiquinone content. These data are congruent with the proposed model that unprotected C-3 hydroxyl of kaempferol triggers the oxidative release of its B-ring as 4-hydroxybenzoate, which in turn is incorporated into ubiquinone. Ubiquinone content in the ugt78d1/ugt78d2 double knockout represented 160% of wild-type level, matching that achieved via exogenous feeding of 4-hydroxybenzoate to wild-type plants. This suggests that 4-hydroxybenzoate is no longer limiting ubiquinone biosynthesis in the ugt78d1/ugt78d2 plants. Evidence is also shown that the glucosylation of 4-hydroxybenzoate as well as the conversion of the immediate precursor of kaempferol, dihydrokaempferol, into dihydroquercetin do not compete with ubiquinone biosynthesis in A. thaliana.