Ammonium sulfate, as well as potassium phosphate, can be used to measure solubility differences between hemoglobin S and hemoglobin A. In addation, the solubility of deoxyhemoglobin C(Harlem) in 1.96 M phosphate has a markedly different temperature dependence from that of deoxyhemoglobin S. This observation indicates that the solubility measurement is quite sensitive to changes in protein structure. Because of the large degree of comparability between the solubility and the aggregation of deoxyhemoglobin S, solubility was used to measure the effectiveness of organic compounds as noncovalent modifiers of deoxyhemoglobin S aggregation. Organic solvents (ethanol, dimethylsulfoxide, 1,4-dioxane, dimethylformamide) alter the solubility characteristics of deoxyhemoglobin S in 1.96 M phosphate buffer, pH 7.0. The concentrations of solvent necessary to provide a half-maximal effect are remarkably similar (about 0.5 M), although the chemical nature of these compounds is quite different. The effect of these solvents must be to prevent the noncovalent bond formation necessary to produce the insoluble hemoglobin precipitate, perhaps by altering the water structure around the deoxyhemoglobin S molecules. In addition to these organic solvents, guanidine hydrochloride and urea, two well-known protein denaturants, were studied. Guanidine hydrochloride was as effective as the best organic solvent in increasing the solubility of deoxyhemoglobin S; urea was far less effective. Studies in vitro with intact erythrocytes from individuals homozygous for hemoglobin S showed that sickling is decreased up to 50% by treatment with ethanol. This offers further evidence that solubility is monitoring a phenomenon similar to the aggregation of deoxyhemoglobin S inside erythrocytes. While use of these particular compounds in vitro would seem to have no clinical implications, these studies do suggest that the use of chemicals that do not modify hemoglobin S covalently should be explored in efforts to prevent deoxyhemoglobin S aggregation.