An analytical system developed for fractionating free and monoacetylated polyamines [Hiramatsu, K. et al. (1994) J. Biochem. 115, 584–589] was proved useful also in detecting diacetylpolyamines, namely N¹,N⁸-diacetylspermidine (diAcSpm) and N¹,N¹²-diacetyl-spermine (diAcSpm). Detection limits were 0.9 and 0.6pmol (S/N=5) for diAcSpd and diAcSpm, respectively. Analytical recovery and within-run variation were also satisfactory. Human urine samples were found to contain diAcSpd and diAcSpm. These polyamines were identified on the basis of the following observations: (i) their retention times were coincident with those of authentic samples; (ii) they were deacetylated to N⁸-acetylsper-midine and monoacetyl- and free spermine, respectively, by acetylpolyamine amidohy-drolase; and (iii) they were practically inert to direct oxidation by bacterial polyamine oxidase as were authentic samples. The amounts of eleven polyamine species including diAcSpd and diAcSpm in urine samples from 52 healthy persons were determined. Mean values for the major polyamine components were consistent with those reported by others. Although the amounts of diAcSpd and diAcSpm were very small, comprising only 1.4 and 0.46% of total polyamines, respectively, these two compounds were found to be always present in healthy human urine as regular constituents. Moreover, variation in their content among individuals was small, suggesting that excretion of these components in urine is strictly regulated.