Quantitative changes of 419 gene transcripts during postnatal mouse cerebellar development were accurately determined with a novel polymerase chain reaction (PCR)-based technique. About 70% of the genes showed differences in expression levels, and the magnitude of difference was relatively small. By hierarchic cluster analysis of developmental expression patterns, the genes were categorized into 19 clusters, which were subsequently assembled into four major groups: group 1, with elevation of gene expression throughout the time course; group 2, with relatively unchanged levels; group 3, with transiently high expression at ∼ 12 days; and group 4, with highest expression at ∼ 4 days. Genes related to brain functions were segregated into several clusters of group 1 and group 3: the same clusters in which cerebellum-specific genes were also segregated. Genes for protein synthesis belonged to group 4. Genes with housekeeping functions belonged to group 2. Western blotting analysis of representative protein products of each group revealed correlation with the mRNA level for those belonging to group 1 and group 4, but not necessarily in the other groups. The close correlation of algorithmically categorized temporal expression patterns of genes with their functions will be useful for estimating the functions of thousands of novel genes.