It is estimated that about half of all therapeutic agents are chiral, but most of these drugs are administered in the form of the racemic mixture, i.e. a 50/50 mixture of its enantiomers. However, chirality is one of the main features of biology, and many of the processes essential for life are stereoselective, implying that two enantiomers may work differently from each other in a physiological environment. Thus, receptors or metabolizing enzymes would recognize one of the ligand enantiomers in favour of the other. With one exception, all presently marketed proton pump inhibitors (PPIs)--omeprazole, lansoprazole, pantoprazole and rabeprazole--used for the treatment of gastric acid-related diseases are racemic mixtures. The exception is esomeprazole, the S-enantiomer of omeprazole, which is the only PPI developed as a single enantiomer drug. The development of esomeprazole (an alkaline salt thereof, e.g. magnesium or sodium) was based on unique metabolic properties that clearly differentiated esomeprazole from omeprazole, the racemate. At comparable doses, these properties led to several clinical advantages, for example higher bioavailability in the majority of patients, i.e. the extensive metabolizers (EMs; 97% in Caucasian and 80-85% in Asian populations), lower exposure in poor metabolizers (PMs; 3% in Caucasian and 15-20% in Asian populations) and lower interindividual variation. For the other, i.e. racemic, PPIs there are some data available on the characteristics of the individual enantiomers, and we have therefore undertaken to analyse the current literature with the purpose of evaluating the potential benefits of developing single enantiomer drugs from lansoprazole, pantoprazole and rabeprazole. For lansoprazole, the plasma concentrations of the S-enantiomer are lower than those of the R-enantiomer in both EMs and PMs, and, consequently, the variability in the population or between EMs and PMs is not likely to decrease with either of the lansoprazole enantiomers. Furthermore, plasma protein binding differs between the two lansoprazole enantiomers, in that the amount of the free S-enantiomer is two-fold higher than that of the R-enantiomer. This will counteract the difference seen in total plasma concentrations of the enantiomers. Also, studies using expressed human cytochrome P450 isoenzymes show that the metabolism of one enantiomer is significantly affected by the presence of the other, which is likely to result in different pharmacokinetics when administering a single enantiomer. For pantoprazole, there is a negligible difference in plasma concentrations between the two enantiomers in EMs, while the difference is substantial in PMs. The difference in AUC between PMs and EMs would decrease to some extent, but in the majority of the population the variability and efficacy would not be altered with a single enantiomer of pantoprazole. The metabolism of the enantiomers of rabeprazole displays stereoselectivity comparable to that of lansoprazole, i.e. the exposure of the R-enantiomer is higher than that of the S-enantiomer in EMs as well as in PMs, which, by analogy to lansoprazole, makes them less suitable for development of a single enantiomer drug. Furthermore, the chiral stability of the rabeprazole enantiomers may be an issue because of significant degradation of rabeprazole to its sulfide analogue, which is subject to non-stereoselective metabolic regeneration of a mixture of the two enantiomers. In conclusion, in contrast to esomeprazole, the S-enantiomer of omeprazole, minimal if any clinical advantages would be expected in developing any of the enantiomers of lansoprazole, pantoprazole, or rabeprazole as compared with their racemates.