A rice (Oryza sativa) T-DNA insertion population, which included more than 63 000 independent transgenic lines and 8 840 identified flanking sequence tags (FSTs) that were mapped onto the rice genome, was developed to systemically study the rice seed quality control. Genome-wide analysis of the FST distribution showed that T-DNA insertions were positively correlated with expressed genes, but negatively with transposable elements and small RNAs. In addition, the recovered T-DNAs were preferentially located at the untranslated region of the expressed genes. More than 11 000 putative homozygous lines were obtained through multi-generations of planting and resistance screening, and measurement of seed quality of around half of them, including the contents of starch, amylose, protein and fat, with a nondestructive near-infrared spectroscopy method, identified 551 mutants with unique or multiple altered parameters of seed quality. Analysis of the corresponding FSTs showed that genes participating in diverse functions, including metabolic processes and transcriptional regulation, were involved, indicating that seed quality is regulated by a complex network.