Many Studies have demonstrated the functions of individual genes associated with embryogenesis and have determined the genome sequences of several organisms. Despite the availability of enormous amount of genetic information, dynamic changes that occur during embryogenesis have not yet been completely understood. In order to understand the dynamic processes involved in embryogenesis, we employed the metabolomic approach. The results of our study indicated that there is a close correlation between metabolomes and developmental stages. Our method enables the identification of embryonic stages using metabolomes as "fingerprints." In this manner, we could successfully predict embryonic development on the basis of metabolomic fingerprints. This is the first report describing a model for predicting vertebrate development by using metabolomics. (C) 2009 Elsevier Inc. All rights reserved.