Rice, the staple food for more than half of humankind and the model monocot crop, was the first cereal in which efficient gene transfer procedures were implemented: 30 years ago, the first transgenic rice plant was regenerated following direct gene transfer to cell suspension-derived protoplasts. Shortly thereafter, transgenic plants were regenerated from zygotic embryo-derived cells following their subjection to micro-projectile acceleration and Agrobacterium-mediated transfection treatments. The high efficiency of transfer deoxyribonucleic acid (T-DNA) integration in seed embryo-derived cells of rice has allowed the transfer of genes of agronomical relevance and the generation of large collections of insertion lines and has provided a key contribution to deciphering the function of more than2000 rice genes. The high efficiency of T-DNA integration in seed embryo-derived cells of rice also permitted the first implementation of gene targeting and knock-in(KI) events, relying on the albeit very low natural frequency of homology-directed pair (HDR) in the rice genome. In the late 2000s, the advent of site-directed nucleases (SDNs) that induce either single or double-strand breaks at a high frequency and their rapid application to rice permitted routine targeted mutagenesis, which can be multiplexed to simultaneously alter several targets or create deletions, and base and gene editing (e.g. correction of amino acids). Currently, the challenge remains to attain a high frequency of KI and replacement of long stretches of DNA for protein domain or coding sequence swapping. We present herein a historical perspective of the advances that have been readily implemented to determine the function of rice genes and to manipulate traits of agronomical relevance. Two main bottlenecks remain to be alleviated in rice genomic engineering: the low frequency of HDR and the genotype dependence of gene transfer efficiency. Alleviation of these bottlenecks is needed to reach the potential