A study of the pH of saliva produced by humans sucking hard candy lozenges containing zinc gluconate and citric acid demonstrated the probability that the formulation delivered an insignificant amount of the contained zinc as Zn2+. This could account for the negative results of several clinical studies of this lozenge and similar formulations as treatment for the common cold. Direct measurement of unbound Zn2+ in saliva from this and other zinc gluconate formulations was required to substantiate the inference from the pH study. A specific-ion-electrode assay method was developed. Using this method, it was found that saliva completely suppresses the ionization of zinc to free Zn2+ in the presence of citric acid or a 30-fold molar excess of mannitol plus sorbitol. Under the same conditions, however, the presence of an excess of glycine does not interfere with ionization to produce Zn2+. This finding supports the hypothesis that the positive clinical results of three studies were due to the use of formulations which release ionic zinc.