Abstract
Computer simulation uses two experimental HPLC runs to allow prediction of sample retention as a function of mobile phase composition or gradient conditions. The general approach is rigorous, but it is assumed that reversed-phase retention obeys the relationship log k' = log kw - S 0. Small deviations in this relationship can lead to error in predicted retention times. We have studied this error empirically for several different reversed-phase systems. This provides a basis for partially correcting these errors, and suggests recommendations for avoiding significant errors during computer simulation.