The number of binding sites for a given solute in a protein is a most relevant parameter. This number can be derived from fluorescence quenching data which provides the fraction of sites occupied at a given free solute concentration. Data are generally treated according to Scatchard´s or Ward´s equations. Lately, a double logarithmic plot of the data has been extensively used with this purpose. The present communication discus the validity of this procedure. It is concluded that this type of plot provides an evaluation of the stoichiometry (molecularity) of the binding process but not the number of equivalent binding sites per protein.