An M13 bacteriophage-based color sensor, which can change its structural color upon interaction with a gaseous molecule, was evaluated as a screening tool for the discrimination of the geographical origins of three different agricultural products (garlic, onion, and perilla). Exposure of the color sensor to sample odors induced the self-assembled M13 bacteriophage bundles to swell by the interaction of amino acid residues (repeating units of four glutamates) on the bacteriophage with the odor components, resulting in a change in the structural color of the sensor. When the sensor was exposed to the odors of garlic and onion samples, the RGB color changes were considerable because of the strong interactions of the odor components such as disulfides with the glutamate residues on the sensor. Although the patterns of the color variations were generally similar between the domestic and imported samples, some degrees of dissimilarities in their intensities were also observed. Although the magnitude of color change decreased for perilla, the color change patterns between the two groups were somewhat different. With the acquired RGB data, a support vector machine was employed to distinguish the domestic and imported samples, and the resulting accuracies in the measurements of garlic, onion, and perilla samples were 94.1, 88.7, and 91.6%, respectively. The differences in the concentrations of the odor components between both groups and/or the presence of specific components exclusively in the odor of one group allowed the color sensor-based discrimination. The demonstrated color sensor was thus shown to be a potentially versatile and simple as an on-site screening tool. Strategies able to further improve the sensor performance were also discussed.