Previous studies showed evidence that quercetin can bind DNA by intercalation [R. Solimani et al., J. Agric. Food Chem. 43 (1995) 876-882] and a comparison with the flavanol dihydroquercetin indicated that the interaction is correlated to the planarity and hydrophobicity of the benzopyranic-4-one plane [R. Solimani, Int. J. Biol. Macromol. 18 (1996) 287-295]. In this study flow linear dichroism (LD) spectra of the hydrophobic quercetin were compared with hydrophilic aglycoside morin and 3-glycoside rutin in the same conditions: [DNA] = 3.1 x 10(-2) mol/l phosphate, [dye] = (1.0-4.0) x 10(-4) mol/l. Morin and rutin in an aqueous environment showed the same behaviour as quercetin in buffer-ethanol (70:30, v/v) solution, with their common benzopyranic-4-one part within the biopolymer. The LD(R) values (LD normalised to the relative isotropic absorption) indicated a greater affinity of the quercetin for the DNA. Comparison of the LD(R) of morin and rutin showed a ratio LD(R)morin/LD(R)rutin approximately 1.1-1.2 very close to unity and this suggests the localisation of the 3-rutinoside of rutin outside the intercalation site. Dichroic measurements recorded in extreme conditions of concentration partly clarified the sequences of interaction between quercetin and DNA in solution which shows the prototypical behaviour of the flavonolic group. This consists of an initial weak external association, where an electrostatic component is excluded, and which can evolve to intercalation changing the DNA concentration, whereas the quantity of the flavonol influences relatively the association. The carbonylic region of the benzopyranic-4-one chromophore is probably localised outside the intercalation site. This was suggested by indirect infrared (attenuated total reflection ATR) data of the quercetin-ethanol solution: the presence of free and chelated carbonyl determines a greater density of negative charges in this region of the chromophore, with the consequent lower probability of this portion penetrating the external polyanionic perimeter of the DNA. A simple approach to determine the order of magnitude of the anisotropic band II of the flavonols completely covered by the more intense DNA band at 260 nm, was also proposed. The low number of intercalated chromophores did not determine an alteration of the flexibility and hydrodynamic behaviour of the biopolymer and this can be correlated to a biological consideration: the flavonols probably do not interfere with the genetic functionality of the DNA. In contrast, the potentially close relationship between these nucleophilic dyes and the biopolymer, shown in this study, suggests a protective role on the nucleophilic groups of the DNA, which are a target of free radicals and the reactive electrophilic groups.