A new procedure to accurately measure a trace amount of D-proline in biological samples has been developed. This D-amino acid was derivatized with 4-fluoro-7-nitro-2,1,3-benzoxadiazole and was determined by a column-switching HPLC system, a combination of a micro-ODS column and a chiral column. The detection limit for D-proline spiked in a mouse cerebrum sample is 1 fmol (injection amount, S/N โซุโฌ 3). Within-day precision and day-to-day precision obtained for spiked D-proline (10 fmol) are 2.14 and 5.35% (RSD), respectively. Using the new method, the amount of free D-pro- line in eight brain regions and sera of mutant ddY/DAO ุ mice, lacking D-amino acid oxidase activity, and control ddY/DAO ุ mice was determined. The amount of free D-leucine was also investigated. The amount and distribution of D-proline in the brains of ddY/DAO ุ mice and ddY/DAO ุ mice are almost the same, and relatively high amounts of D-proline have been observed in the pituitary gland and in the pineal gland. On the other hand, the amount of D-leucine is different between the two strains. In the brains of ddY/DAO ุ mice, a relatively high amount of D-leucine has been observed in the pineal gland compared with other regions. In the brains of ddY/DAO ุ mice, D-leucine amounts are approximately 10 times higher than those obtained in ddY/DAO ุ mice and regional difference has not been observed, while the amounts of L-proline and L-leucine are not significantly different between the two strains. In the serum, the amounts of both free D-proline and D-leucine are significantly higher in the ddY/DAO ุ mice than those obtained in ddY/DAO ุ mice.