The design and synthesis of metal-organic frameworks (MOFs) have evolved into a rich interplay of chemistry principles, topology engineering, and now data-driven methods. By leveraging fundamental concepts like coordination preferences and connectivity, researchers can rationally pre-design framework structures on paper and then realize them in the lab—a far cry from the purely serendipitous discovery of porous solids in the past. Synthetic techniques such as modulated self-assembly and post-synthetic modification provide fine control over crystallinity, porosity, and functionality, allowing the creation of MOFs tailored to specific tasks. Meanwhile, the rise of machine learning in this field is opening new avenues to handle the complexity of MOF chemistry, offering predictive insights and novel design suggestions that can greatly reduce the trial-and-error cycle. As our libraries of MOF building blocks and known nets expand, so too does the confidence in constructing ever more complex frameworks (e.g., multi-SBU MOFs, hierarchical or interwoven nets by design, biomimetic MOFs with enzymatic functions). By continuing to refine our understanding and harness new technologies, MOF researchers are steadily driving the creation of materials perfectly suited for the challenges ahead.