While neutral and acidic compounds are well separated by upercritical fluid chromatography (SFC), basic analytes are more challenging to separate and often problems occur with their peak shapes. Two different methods were explored in the present paper to reduce these problems and maintain compatibility with mass spectrometry (MS). Five different, commercially available 2-ethylpyridine (2-EP) stationary phases were tested without a mobile phase additive using 92 pharmaceutical compounds with basic properties. The kinetic performances of the 5 columns were nearly identical, but the peak shapes of the basic drugs were strongly affected by the stationary phase. The PrincetonSFC 2-EP and Zymor Pegasus 2-EP phases clearly outperformed the other stationary phases, with 77% and 69% of the compounds having Gaussian peaks (and asymmetries between 0.8 and 1.4), respectively. Comparatively, the Waters Viridis Silica 2- EP, Waters Viridis BEH 2-EP and ES industries GreenSep 2-EP phases provided only 52%, 44% and 22% of the compounds with Gaussian peaks, respectively. These differences were attributed to the significant dissimilarities in their silica matrix properties. An alternative strategy was also performed with a hybrid silica stationary phase, Viridis BEH, using 20 mM ammonium hydroxide in the mobile phase, which was a mixture of CO2 and MeOH. With these conditions, 81% of the peaks observed for the basic analytes were Gaussian; however, this value dropped to 17% and 10% in the absence of additive and in the presence of 20 mM formic acid, respectively. Finally, the use of a hybrid bare silica stationary phase in the presence of 20 mM ammonium hydroxide is quite an interesting solution as this system is compatible with both ultra high performance SFC (UHPSFC) columns packed with sub-2 m particles and with MS detection. The overall applicability of this system was demonstrated with various mixtures of basic drugs.