In this work, we modelled, fabricated and fully characterized a multiplexed fiber Bragg grating sensor embedded in a soft 3D-printed flexible skin to evaluate the finger joint movements of the hand, with a specific focus of this work on the index finger. Bragg gratings with separation between each other defined according to the finger’s phalanges were used for that. These were embedded in a 1 mm thick elastic 3D printed membrane, allowing the FBG to strain and respond to curvature. Curvature characterizations were implemented through two different methodologies, namely through the movement of one terminal related to the other (inefficient), and through the use of 3D printed parts with predefined curvatures. The later method revealed better calibration performances. As a proof of concept, a 3D-printed mold with predefined curvatures, emulating the index finger flexure was fabricated and the response of the embedded sensor was evaluated. The results showed the capability to measure the joint movement with a maximum error of 3.4 mm in radius of curvature (the maximum difference between the experimental values and the actual values), showing its ability to be used in a real-world scenario. This work paves the way for future flexible wearable sensors capable of aiding occupational therapists in the evaluation of the patient’s hand mobility, allowing them to properly define the training program and thus, enhancing the quality of life of persons with hand disabilities.