Potentiometric titration curves have traditionally been collected as the difference in absorbance at two wavelengths, and analyzed by plotting voltage vs. log (oxidized/reduced). The collection method, designed to monitor changes in local peak height, is effective for that purpose only when spectral backgrounds do not change slope as voltage changes, and the analysis method is valid only for a single isolated component (one whose midpoint potential is far from that of anything else in the mixture). Yet these methods are commonly used where such restrictions do not pertain, e.g. the study of cytochromes in mitonchondria. In this paper, we present more appropriate methods of collection and analysis, and suggest that, even with the best available methods, any conclusion should be confirmed in several ways. Experimental results are presented in accompanying papers.