Of the 20 and the 28 compounds required to reproduce the compositions of (1) the cleavage mixture in the acetone distillation unit and of (H) the cyclohexanot-cyclohexanone mixture in the Polish process for making the two compounds, 3 and 7 were nonexistent in the ChemCad Data Bank, resp. With UNIQUAC selected to describe phase equilibria, (i) 210 and (ii) 378 binary interaction parameters (BIP) were required, of which only (i) 20 and (ii) 23 were available. The missing BIP were predicted. A drastic fallacy was the hexanal-cyclohexanone system for which UNIFAC predicted positive, and the experiment showed considerable negative deviations to occur from ideallty. This single mispredicted equilibrium misled the simulations considerably. With the experimental data used, the reflux had to be doubled and the forerun was 2.5 times as big as the figures obtained with the predicted equilibrium data.