Taking gold nanoparticles with different hydrophilic coatings as an example, it is investigated whether capillary electrophoresis in combination with Taylor dispersion analysis allows for the precise determination of mean electrophoretic mobilities, electrophoretic mobility distributions, and zeta potentials in a matrix of exactly known composition and the calibration-free determination of number-weighted mean hydrodynamic radii. Our experimental data confirm that the calculation of the zeta potential for colloidal nanoparticles with zeta > 25 mV requires to take the relaxation effect into account. Because of the requirement to avoid particle wall interactions, a solution of disodiumtetraborate decahydrate (borax) in deionized water had been selected as suitable electrolyte. Measurements of the electrophoretic mobility at different ionic strength and application of the analytic approximation developed by Ohshima show that in the present case of a buffered solution with a weak electrolyte co-ion and a strong electrolyte counterion, the effective ionic drag coefficient should be approximated with the ionic drag coefficient of the counterion. The obtained results are in good agreement with theoretical expectations regarding the dependence of the zeta potential and the electrokinetic surface charge density on the ionic strength. We also show that Taylor dispersion analysis (besides estimation of the number-weighted mean hydrodynamic radius) provides additional information on the type and width of the number-weighted particle distribution. (C) 2015 Elsevier Inc. All rights reserved.