Accurate limiting activity coefficients can be obtained from differential ebulliometry if vapor and liquid holdups in various parts of the equipment are correctly accounted for in calculating the equilibrium liquid composition from the known charge composition. Available procedures require the difficult-to-determine dynamic vaporization rates (V/F) or (V/L). Two new procedures proposed determine V/F (= phi): 1. a trial-and-error procedure satisfying all three of the equations governing the ebulliometer operation yielding the true liquid equilibrium composition x(1), phi, and the overall composition in the Cottrell pump, Z(1); 2. one procedure based on the measured superheat temperature at the base of the Cottrell pump. A possibly limiting assumption is that local equilibrium exists at this point, but the method has good rigour otherwise. It is, however, very sensitive to measured pressures and temperatures, which need to be of the highest accuracy. The procedure is illustrated for the n-hexane (1)-ethanol (2) and n-hexane (1)-ethylacetate (2) systems, and requires liquid density, heat capacity and accurate latent heats over a small temperature interval. Measurements suggest that about 72% of the total vaporization takes place at or before the entrance to the Cottrell pump with only about 28% in the equilibrium chamber for the n-hexane-ethyl acetate system. For the n-hexane-ethyl alcohol systems, they are 89% and 11%. The iterative procedure is recommended because it is more reliable and, although computationally extensive, has more rigor.