During animal development, cells are genetically controlled to generate the three-dimensional morphological structures. A collection of quantitative data of spatiotemporal morphological dynamics when a wide variety of individual genes are perturbed would provide a rich resource for understanding animal development. Here we created a collection of quantitative data of cell division dynamics in early C. elegans embryos when each of all 351 essential embryonic genes was silenced individually by RNA interference (RNAi). To collect five sets of quantitative data from RNAi-treated embryos for each gene, we applied our computer image processing system to ~2,000 sets of four-dimensional differential interference contrast microscopy images. The collection finally consists of 33 sets of quantitative data for wild-type embryo and 1,142 sets of quantitative data for RNAi-treated embryos for 243 genes. Each data contains the 3D coordinate values of the outlines of nuclear regions and the changes of the outlines over time. We first performed computational phenotype analysis by using this collection. Statistical analysis identified over 6,000 RNAi-induced phenotypes for 437 phenotypic characters, which can be numerically extracted from the collection. The phenotypes include three-dimensional (3D) spatial and temporal phenotypes that are difficult to be identified manually. We next predicted sets of genes involved in the same cellular processes from the collection. By applying hierarchical clustering method to the profiles of the extracted phenotypic characters, we found 7 independent clusters corresponding to the individual cellular processes such as polarity/asymmetric division, DNA replication and chromosome maintenance/segregation. In addition, we developed novel computational methods for inferring a coarse-grained model of embryogenesis by finding correlations between phenotypic characters. The inferred model represents the order in which the phenotypic characters relate and appear in the course of development. Our collection provides novel opportunities for performing high-throughput analysis of 3D spatial and temporal cell division dynamics during animal development. In the near future, all data and results will be available at the Worm Developmental Dynamics Database (http://so.qbic.riken.jp/wddd).