The contemporary urban condition is one of constant change and flux, and cities, like complex organisms, have evolved over time to meet the needs of these changing conditions. Traditional city planning attempts to predict, control, or subvert these changes, often with mixed results. Drawing inspiration from the natural processes of growth and evolution, this project proposes the use of algorithmic design to inform the planning and growth of future cities. “Generating the Image of the City” is an experiment in multi-agent code that attempts to generate city form through a simulated urban embryogenesis, taking as its DNA Kevin Lynch’s five elements of the image of the city. By varying the behavior and interactions between districts, edges, landmarks, nodes and paths, the designer is able to exert a certain degree of control over the ultimate form.