No comparative information is available concerning the ability of various cholinesterase (ChE) methods to identify succinyldicholine-sensitive patients, purely on the basis of the enzyme activity recorded in serum. Here, we evaluated six different methods for the measurement of ChE activity; 131 subjects were subdivided according to ChE phenotype and, therefore, to succinyldicholine sensitivity. ChE phenotype was determined by measuring dibucaine and fluoride numbers. DNA analysis was also performed to confirm correlation between the phenotype classification used in the study and the ChE genotype. The tested methods were significantly different in their ability to discriminate between the subjects with and without succinyldicholine-sensitive phenotypes. The succinyldithiocholine/5,5'-dithio-bis(2-nitrobenzoate) (DTNB) method showed the highest accuracy (area under the receiver operating characteristic (ROC) curve 0.97) followed by the propionylthiocholine/DTNB method (area under the ROC curve 0.94). On the other hand, the two methods using butyrylthiocholine as substrate and that employing benzoylcholine showed limited clinical utility in discriminating subjects at risk of prolonged apnea (area under the ROC curve < or = 0.9). Using the succinyldithiocholine method, a value < or = 23 U/l was approximately five times as likely to occur in a sensitive individual as in a normal one.