The hypernetted chain (HNC) theory is used to compare charge titration and zeta potential data for several colloids. It is found that the HNC diffuse layer potential is closer to the zeta potential than that calculated with the traditional Poisson-Boltzmann (PB) approximation, and hence the amount of counterion binding required to bring the two into conformity is reduced. Further improvement is obtained by using an effective PB diffuse layer potential, calculated from both the numerical HNC theory and the analytic extended PB approximation. The rationale for using the effective PB potential is that the conversion from the measured electrophoretic mobility to the zeta potential invokes the PB approximation. In the case of silver iodide quantitative agreement is obtained without any fitting parameters, which confirms that no counterion binding occurs. For the metal oxides analyzed (titanium dioxide, aluminum oxide, and silicon dioxide) the two measurements can only be reconciled by postulating counterion binding or surface charge mobility.