Finger tapping assessment helps clinicians to evaluate the upper limb motor function and motor impairment in patients with neurodegenerative diseases like Parkinson’s and Alzheimer’s. This paper proposes a finger tapping sensor based on the embedding of fiber Bragg grating (FBG) sensors in soft compounds. The embedding procedure demonstrates to enhance the sensitivity of the FBG to the finger pressure and, at the same time, provides flexibility and protection to the sensing element. The proposed device seeks to be a novel non-invasive alternative solution to the sensing systems reported in literature, often complex and bulky, and to support the current clinical assessment, usually qualitative and performed visually. The patch has been tested on four healthy subjects carrying out three kinds of task, i.e., random, rhythmic, and fast index finger tapping, each one executed for 20 seconds and by using both hands, one at a time. The proposed sensor proves to be able to record the temporal evolution of the tapping signals and allows to discriminate the tapping waveform between the different involved subjects and performed tasks, enabling comparative analyses among them.