Functional near infrared spectroscopy (fNIRS) is a safe and portable brain imaging modality that monitors changes in the hemodynamic activity at the cortical level. Although still in its emerging stage, fNIRS has recently gained increasing acknowledgements of its strengths and suitability for many clinical applications. The fast evolution and growth of fNIRS applications has been made possible mainly by studies that substantiate the general validity of the fNIRS measures. Such studies investigate both the fNIRS construct, by cross-validating it with fMRI, and the repeatability of fNIRS measures.Nonetheless, cases exist that would pose a challenge forfNIRS measures of cortical activation. In particular, violations of the assumptions made on the optical properties of the sampled tissue would affect some variables included in the modified Beer-Lambert law (mBBL), which allows conversion of the changes in measured light intensity into changes in the oxyhemoglobin and deoxyhemoglobin concentrations. These violations would therefore reflect on the fNIRS readings and on the way data are interpreted. The aim of this paper is to present an example of such challenging situations. The case presented is a subject whose left frontal lobe cortex has been partially ablated following a subdural hematoma. fNIRS measures were recorded during a verbal fluency task, known to be associated to functioning of the left frontal lobe. We examine the outcome of fNIRS, contextualizing it in the framework of the mBLL and its assumptions.