The role of negatively charged residues in tuning the absorbance spectrum of bovine rhodopsin has been tested by mutating each aspartate and glutamate to asparagine and glutamine, respectively. Previous work demonstrated that aspartate83, glutamate122, and glutamate134 can be replaced by neutral residues with little or no effect on the absorbance spectrum of the resulting pigment [Nathans, J. (1990) Biochemistry 29, 937-942]. With one exception, mutations at the remaining 19 aspartate and glutamate residues result in very nearly wild-type absorbance spectra. The exception is glutamate113: mutation to glutamine causes the pigment to absorb at 380 nm, reflecting deprotonation of the retinylidene Schiff's base. Upon addition of either chloride, bromide, or iodide, the absorbance rapidly shifts to 495, 498, or 504.5 nm, respectively, reflecting protonation of the Schiff's base. The progressive red shift observed upon addition of halides with larger atomic radii strongly suggests that halides are serving as the Schiff's base counterion. Halides have no effect on the absorbance spectrum of wild-type rhodopsin. I infer, therefore, that glutamate113 is the retinylidene Schiff's base counterion in wild-type rhodopsin. Sakmar et al. [(1989) Proc. Natl. Acad. Sci. U.S.A. 86, 8309-8313] and Zhukovsky and Oprian [(1989) Science 246, 928-930] have arrived at the same conclusion based upon a related series of experiments. These data support a model in which spectral tuning in bovine rhodopsin results from interactions between the polyene chain of 11-cis-retinal and uncharged amino acids in the binding pocket.