Two methods are used to establish the influence of concurrent retention mechanisms on the accuracy of McReynolds phase constants for the stationary phases squalane, OV-225, Carbowax 20M, OV-275, ethylammonium nitrate, n-propylammonium nitrate, n-butylammonium thiocyanate, sec.-butylammonium thiocyanate, and di-n-propylammonium thiocyanate. Phase loading studies of the net retention volume per gram of packing as a function of the percent liquid loading were used to indicate, qualitatively, those cases for which adsorption at the liquid and support interfaces made a significant contribution to the retention mechanism of test solutes and the n-alkane retention index markers. It is shown that for some cases accurate phase constants can be calculated from the gas—liquid partition coefficient of the test probes and n-alkanes obtained by extrapolation to infinite phase volume of plots of VN/VL against 1/VL (VN = net retention volume and VL = volume of liquid phase). This approach was successful for obtaining gas-liquid partition coefficients for all polar probes but not always for the n-alkane retention index markers. The n-alkanes were retained predominantly by interfacial adsorption on OV-275, ethylammonium nitrate, and n-propylammonium nitrate. In these cases accurate McReynolds phase constants can not be determined by any method that relies on measurements based on the partitioning of the n-alkanes. Alternatively, the relative selectivity of all phases evaluated can be determined from the partial molar free energy of solution for the McReynolds test probes. Infinite dilution activity coefficients and the partial molar free energy of solution for a methylene group are used to help establish the nature of solution interactions between the n-alkanes, McReynolds test probes, and the phases studied.