The design of plants or unit operations requires the knowledge of data for multicomponent systems. An approach is presented for the regression of binary interaction parameters required for the calculation of activity coefficients. Binary vapor pressure data are used to correlate these parameters for calculation of vapour/liquid equilibrium data. The advantage of the approach is the fact that only easily measured vapor pressure data are required. The procedure is demonstrated by means of two binary mixtures. Parameter regression, prediction of vapour/liquid equilibrium data as well as their experimental verification are demonstrated.