Yamazaki' study and a study by Yamazaki and Eto indicated that a combination of black text and a background color with high luminance and high brightness was not considered preferable for Web-based tests (WBTs). In this study, the authors conducted an experiment to examine if the scores of on-line English tests differed depending on the characteristics of a background color. By using near-infrared spectroscopy, relative changes in Hemoblobin (Hb) concentrations in the brains of test takers were observed to see how background colors can affect the functions of their brains. Twenty four subjects in their twenties took web-based English grammar and non-linguistic tests with white and blue background colors with black text. The average scores of the linguistic and non-linguistic tests for the blue background were higher than those for the white background among the subjects. In particular, a significant difference in subjects' performance was found between the white and blue backgrounds for the non-linguistic task. Two dimensional images of the Hb concentration changes obtained in the experiment showed that areas in the brain related to the frontal eye field were observed to be more active while the subjects were taking the tests with a white background. These results indicate that the combination of black text and white background encourages a test-taker to concentrate more on visual input from the screen rather than the test questions. They also suggest that the combination may not be the best choice for a background color of a WBT to assess test-takers' linguistic performance even though a white background is commonly used for WBTs.