This paper presents our efforts in the development of a small wireless, flexible bandage sized near-infrared spectroscopy (NIRS) system for sleep analysis. The current size of the system is 2.8 cm × 1.7 cm × 0.6 cm. It is capable of performing NIRS with 660nm, 940nm and 850nm wavelengths for up to 11 hours continuously. The device is placed on the forehead to measure from the prefrontal cortex and the raw data is continuously streamed over Bluetooth to a nearby data aggregator such as a smartphone for post processing and cloud connection. In this study, we performed traditional polysomnography simultaneously with NIRS to evaluate agreement with traditional measures of sleep and to provide labelled data for future work involving learning algorithms. Ultimately, we expect a machine learning algorithm to be able to generate characterization of sleep states comparable to traditional methods based on this biophotonics data. The system also includes an inertial measurement unit and the features that can be extracted from the presented system include sleep posture, heart rate, respiratory rate, relative change in oxy and deoxy hemoglobin concentrations and tissue oxygenation and cerebral arterial oxygen extracted from these. Preliminary proof of concept results are promising and demonstrate the capability to measure heart rate, respiratory rate and slow-wave-sleep stages. This system serves as a prototype to evaluate the potential of a small bandage-size continuous-wave NIRS device to be a useful means of studying sleep.