Interethnic variation amongst the drug metabolising enzymes relevant to the treatment of psychosis is reviewed. The frequency of genetically determined variants at the extremes of enzyme activity is seen to vary considerably between different ethnic groups; in addition, a shift in the frequency distribution giving an overall lower population mean activity may occur. The role of dietary and other environmental influences in the generation of interethnic variation in cytochrome activity is also discussed. Clinical studies pertinent to this variation are reviewed. It is suggested that the reason for conflicting data from some clinical studies is the existence of overlapping substrate specificity, so that one cytochrome is able to substitute for another. Individuals deficient for more than one cytochrome would be likely to show much more pronounced clinical effects than those showing single cytochrome deficiency.