Four adsorption equations were fitted to phosphate isotherm data for 38 soils from northern New South Wales. The two-surface Langmuir equation provided the best fit to the data. The Freundlich equation, however, was almost as effective, provided that an estimate of native adsorbed phosphate was included. It required fewer adsorption points because the transformed data produced a straight line. A regression constant, a, from the Freundlich equation although underestimating adsorption capacity, calculated from the Langmuir II equation, was highly correlated with this parameter (r = 0.984) and a buffering index (r = 0.986), calculated from the Langmuir I equation. The other Freundlich regression constant, n, was significantly correlated (P <1%) with the intensive parameters of adsorption from the one- and two-surface Langmuir equations. Since only two adsorption points are required for fitting the Freundlich equation and the parameters are easier to derive, it is suggested that this equation may be more suitable for commercial soil-testing laboratories than the Langmuir one- and two-surface equations for routine determination of phosphate adsorption characteristics of soils.