Humans use various higher cognitive processes for taste, including memorization, recollection, and imagination, in our dietary lives, as well as in more specialized situations, such as sensory evaluation used in the food industry. So far, the cognitive aspects of taste processing have been studied mainly with psychological approaches, and their neural bases are not yet well understood. Conventional neuroimaging techniques are being used to converge psychological findings with human brain functions. However, these techniques require subjects to be in a supine position and strictly restrict head movements, narrowing the range of experimental paradigms that can be performed. This is especially true for taste studies; it is difficult to taste samples and perform cognitive tasks on them under conventional neuroimaging conditions. Therefore, we have attempted to use functional near infrared spectroscopy (fNIRS), an emerging noninvasive neuroimaging technique, to study human cortical taste cognitive processing. fNIRS, utilizing light to measure cortical hemoglobin concentration changes associated with neural activities, is more tolerant of subject body movement, thus allowing a wider range of experimental tasks for taste. However, it has some shortcomings that needed to be addressed. In this review, we will show how these technical obstacles have been overcome, how fNIRS contributes to the mapping of taste-related brain functions, and further promote the understanding of human taste processing. We propose fNIRS as a potential mediator between psychology and neuroscience.