Unperturbed dimensions of flexible linear macromolecules can be obtained from [η]-M-data in any solvent, good or poor, single or mixed. Usually Kθ is estimated by a relationship between [η]/MW0.5 and Mw0.5 first proposed by Burchard and by Stockmayer and Fixman. But, it is well-known that the Burchard-Stockmayer-Fixman-plot shows downward curvature, especially for good solvent systems. Various efforts have been made to achieve relations with better linearity. One of the first was the semi-empirical relation between ([η]/Mw0.5)0.5 and Mw/[η] by Berry. Predicting a relationship of the excluded volume parameter z to the viscosity expansion factor by α5η instead of α5η Tanaka obtains that ([η]/Mw0.5)5/3 is linear in Mw0.5. By allowing for the dependence of the viscometric interaction parameter B, which is correlated to the second virial coefficient A2, on molar mass, Gavara, Campos and Figueruelo predict a linear dependence of [η]/Mw0.5 against A2.Mw0.5. It is not our intention here to discuss the validity of these theories, but to compare them with experimental data.