The research team extended its structured public involvement (SPI) protocol to the field of context-sensitive noise wall design. Subject to a minimum sound attenuation capacity, noise walls must present a pleasing visual aspect to their user communities including residents, commuters, and others. This paper details the casewise visual evaluation (CAVE) methodology and discusses its application to a context-sensitive noise wall design in Arizona. The research team designed a fuzzy logic nonlinear modeling process, CAVE, that predicts group preference for specific designs even when such designs have not been created. CAVE offers significant advantages over current visual assessment methods such as the visual preference survey. Key noise wall design parameters, including height, topological variation, color value, plant coverage, and plant complexity, were identified by landscape architecture experts. Group preferences were gathered rapidly, anonymously, and fairly from a focus group by using electronic polling technology to evaluate digital images of samples and potential designs. Highest preference was achieved with berm-type walls combining medium-value, smooth stone featuring a basic pattern with undulating topographic variation and a higher degree of plant complexity. A range of nonlinear preference variations in response to changes in value and plant complexity were noted. Other preferred designs were documented for this context-specific Arizona case study. The advantages of the SPI process using CAVE and obstacles in its implementation are discussed.