This paper presents the design and evaluation of an in-socket sensory system for a transfemoral prosthetic leg using a set of piezoelectric sensors. The design process includes identifying the optimal mounting configuration of the sensors and determining their best placement. Two experiments were performed separately to address each objective. Results of the experiment suggested that cushionall with fit size cantilever with elastic foundation should be adopted to mount the sensors on the socket's wall. As for the placement of the sensors, the result suggested that the sensors should be positioned in zig-zag orientation from top to bottom, which can cover all of the most active area on the quadriceps and hamstring muscle groups. With the identified design, the socket was fabricated and instrumented for a transfemoral amputee and the performance of the in-socket sensory system was evaluated. The amputee performed level walking at normal speed on a 5-m straight pathway multiple times while the output voltage signal from each sensor was recorded. The output signals consistency was assessed by calculating the cross correlation, r, within five trials. In order to measure the strength of the correlation matrix of each sensor, the Frobenius norm, ||A|| _F was calculated. The high norm value for all sensors pattern (||A|| _F >4.5) for quadriceps implies that the mounting and placement of the sensors is most suitable for a consistent and reliable signal. For Hamstring, results inferred that except for the two sensors located on the top most position and the most bottom sensor, the other sensors showed good performance (||A|| _F >4.4). Thus, the proposed design of the in-socket sensory system using piezo sensors was proven to be effective for a transfemoral prosthetic leg.