The approaches and tools for herbicide lead generation and optimization are increasingly sophisticated. Inputs to screens derived from traditional chemistry, natural products, or combinatorial chemistry are tested in vivo against plants, or in vitro against a target enzyme or receptor. In both cases, high throughput capacity is made possible by automated sample handling and data tracking. Specific in vitro targets may be identified and validated by chemical or genetic means. Increasingly, an inhibitor of a target may be designed with the aid of detailed biochemical and computational tools. If the target compound is active, it will be advanced in successive levels of whole plant testing under conditions that increasingly reflect the actual performance needs of the new herbicide. Detailed studies of uptake, translocation, metabolism in the plant, or partitioning in the soil may be used to understand the barriers to activity delivery and guide the optimization effort. Further, activity on nontarget organisms and persistence in the environment are assessed to ensure the lead optimization program is directed toward discovery of products that are safe to the environment and the end user. Although increasingly sophisticated technical tools provide high quality data to help direct decision making, it is the curiosity and initiative of the discovery scientist that is ultimately responsible for a herbicide product breakthrough. Keywords: natural products; combinatorial chemistry; herbicide target sites; high throughput screening; herbicide design; activity delivery; bio-availability; genomics