In recent years, genome editing technologies have advanced from nuclease-based methods that generate DNA double-strand breaks (DSBs), which can cause undesired byproducts, to next-generation CRISPR approaches that perform controlled chemistry using DSB-independent strategies. Base editing and prime editing are ideally suited for small-scale modifications, but methods to achieve large-payload gene insertion have been lacking. To address this gap, I will present work from my lab describing a new family of CRISPR-associated transposases (CAST) that perform highly accurate, targeted integration of DNA payloads via RNA-guided transposition. Unlike conventional CRISPR systems that combine targeting and cleavage, CAST systems exploit nuclease-deficient CRISPR systems for RNA-guided DNA targeting, leading to the site-specific recruitment of transposase enzymes for DNA insertion. We have resolved molecular details of this pathway using a combination of high-throughput sequencing, biochemistry, genetics, and cryo-electron microscopy, revealing a hierarchical assembly pathway and a structural roadmap to guide engineering efforts. More recently, we leveraged phage-assisted continuous evolution (PACE) to identify transposase variants with ~200-fold improved activity in mammalian cells, yielding 10–25% integration efficiencies of kilobase-size DNA cargos across several human genomic sites. Our long-term goal is to harness CAST as a versatile platform technology for integrating large DNA payloads into the genome for basic research and to treat human disease.