There were 716,000 births in the United Kingdom in 2004 of which approximately 10% received support to establish independent breathing. This paper reports the development and characterization of a sensor that will use forehead reflectance photoplethysmography to provide continuous heart rate monitoring of newborns to aid effective resuscitation in the delivery room. The PPG device used modulated light and lock-in detection, utilizing a novel and simple quadrature demodulation algorithm, to reduce the problem of ambient light artifact, and improve signal-to-noise ratio. Green light (500nm-600nm) was used, in preference to red or infrared wavelengths, to optimize the amplitude of the pulsatile signal. Simulation of the lock-in technique and demodulator and calculation of the front-end noise, showed that the sensor noise floor was limited by the shot noise of the detected light carrier. There was a reasonable match between the simulated and measured noise values, but the shot nose limit was not reached. Dark noise was 85.4 nW/m2/?Hz, and light noise was 399 nW/m2?Hz In-vivo measurement of the reflectance PPG signal at the forehead, will be presented that confirm earlier reported results that light between 500 nm and 600 nm has the largest depth of modulation after interaction with pulsatile blood. This paper will also present experimental results, using the above lock-in algorithm, to extract the pulsatile signal from human subjects.