Kielmeyera rubriflora Camb. contains 2-hydroxyxanthone, 2,4-dimethoxy-3-hydroxyxanthone, 2,3-dimethoxy-4-hydroxyxanthone, 4-hydroxy-2,3-methylenedioxyxanthone, 4-methoxy-2,3-methylenedioxyxanthone and 1,7-dimethoxy-2,3,8-trihydroxyxanthone (Ia). Wood and bark of the tree Kielmeyera rubriflora Camb. (Guttiferae) were collected in the Sêrro region, Minas Gerais State, Brasil. The benzene extract of the wood yielded 2-hydroxyxanthone, 2,4-dimethoxy-3-hydroxyxanthone, 2,3-dimethoxy-4-hydroxyxanthone, 4-hydroxy-2,3-methylenedioxyxanthone, 4-methoxy-2,3-methylenedioxyxanthone, kielcorin and β-sitosterol. All the xanthones had been isolated previously from other Kielmeyera species and were identified by direct comparison with authentic samples. The benzene extract of the bark yielded, besides aliphatic material, a new compound. Its molecular weight, determined by mass spectrometry, was consistent with the constitution of a dimethoxytrihydroxyxanthone. The PMR-spectrum confirmed the existence of two methoxyls and of three aromatic protons. One proton was represented by a singlet at 3.55τ and must, consequently, be placed at the 2-, 3- or 4-position of a trioxygenated ring. The remaining two protons were represented by doublets at 2.67 τ (J 9.0 Hz) and 3.43 τ (9.0 Hz) and must, consequently, be placed at the 5,6 or 6,7-positions of the dioxygenated ring. All signals were sufficiently up-field to preclude the existence of a proton at a peri (1,8)-position of either ring. One of these peri-positions is occupied by a hydroxyl, as shown by the shift of the UV absorption maxima upon addition of AlCl₃ + HCl. The second peri-position, however, is occupied by a methoxyl, since the compound reacted speedily with diazomethane to form a monohydroxy-tetramethoxyxanthone (and not a dihydroxy-trimethoxyxanthone). The mass spectrum of this derivative contained a peak corresponding to M-15-18 a.m.u. Loss of the elements of water upon electron impact again suggested the presence of a methoxy-group adjacent to the carbonyl. The di-O-methyl derivative gave a Gibbs test maximum at 667 nm, typical of a system featuring an unsubstituted position, para related to a peri-hydroxyl. The Gibbs test, when performed on the original compound, as well as the shift of UV maxima upon addition of H₃BO₃ + NaOAc, indicated the presence of an ortho-dihydroxy grouping. Clearly, these hydroxyls cannot be situated at positions adjoining the peri-hydroxyl. Three vicinal hydroxy groups are incompatible with the relatively high stability of the compound in presence of alkali. They have, consequently, to be placed either at C-2,C-3 adjoining the peri-methoxyl, or at C-3,C-4. The latter alternative, however, cannot be correct, since it also would impart instability to a xanthone in alkaline medium. Thus, the peri-hydroxyl with the unsubstituted para-position is located in the disubstituted ring. In view of the vicinity of the hydrogens on this ring, the remaining methoxyl and the peri-hydroxyl must be ortho-related. This leads to the constitution of 1,7-dimethoxy-2,3,8-trihydroxyxanthone (Ia) for the new compound. The 1,2,3,7,8-pentaoxygenation pattern has not been reported previously for a natural xanthone.