The Guided Search (GS) model, a quantitative model of visual search, was used to develop menu designs in a four-step process. First, a GS simulation model was defined for a menu search task. Second, model parameters were estimated to provide the best fit between model predictions and experimental data. Third, an optimization algorithm was used to identify the menu design that minimized model predicted search times based on predefined search frequencies of different menu items. Fourth, the design was tested. The results indicate that the GS model has the potential to be part of a system for predicting or automating the design of menus. # 2002 Elsevier Science Ltd.