The excited-state proton-exchange reaction of commonly used fluorescent pH probes at physiological pH becomes reversible upon addition of pH buffer. Using computer-generated fluorescence decay surfaces, we investigated under which experimental conditions (pH, buffer concentration, and excitation and emission wavelengths) the rate constants describing the excited-state processes (k(ij)) and the spectral parameters related to excitation ((b) over tilde (1)) and emission ((c) over tilde (1)) can be accurately and precisely estimated by global compartmental curve fitting. It was found that a minimum of three fluorescence decay traces should be collected for the pH probe in the presence of buffer. These three decays should be characterized by at least two different pH values and at least two different buffer concentrations. In addition to these three traces, a minimum of one trace corresponding to the pH probe without buffer has to be recorded. Furthermore, for the accurate estimation of k(ij), (b) over tilde (1), and (c) over tilde (1), at least two of these traces should be collected at the same pH and excitation and emission wavelengths. The experimental conditions should be chosen in such a way that decays with unambiguous biexponential character are obtained. For fluorescent pH probes with pK(a) ≈ 7 that are responsive in the near-neutral pH range, it is advisable to use buffers with PKaB values comparable to or higher than the pK(a) of the probe. Because the changes in the decay times are already apparent with a small quantity of buffer, there is no need to use excessively high buffer concentrations. From a practical point of view, the best experimental design is attained when one combines in a single fluorescence decay surface traces originating from samples characterized by different pH values at the same buffer concentration with traces characterized by different buffer concentrations at the same pH and decays of samples without buffer measured at several pH values.