The capillary electrophoretic separation of noncharged enantiomers with single-isomer anionic resolving agents is reexamined here with the help of the charged resolving agent migration model. Two general model parameters have been identified that influence the effective mobility, separation selectivity and mobility difference curves of the enantiomers: parameter b, Galled binding selectivity (K-RCD/K-SCD), and parameter s, called size selectivity (mu(RCD)(O)/mu(SCD)(O)) Analysis of the model in terms of these parameters indicates that in addition to the known, previously observed separation selectivity vs. resolving agent concentration patterns, a new pattern, increasing separation selectivity with increasing resolving agent concentration, is also possible provided that (i) K-RCD/K-SCD<1 and (mu(RCD)(O)/mu(SCD)(O)>1 and (K(RCD)mu(RCD)(O))/(K(SCD)mu(SCD)(O))>1, Or (ii) K-RCD/ K-SCD>1 and mu(RCD)(O)/mu(SCD)(O)<1 and (K(RCD)mu(RCD)(O))/(K(SCD)mu(SCD)(O))<1. This hitherto unseen separation selectivity pattern was experimentally verified during the capillary electrophoretic separation of the enantiomers of O-isapropyl p-nitrophenyl methylphosphonate with the single-isomer octakis(2,3-diacetyl-6-sulfato)-gamma-cyclodextrin as resolving agent.