Although epidermal melanin content has been quantified non-invasively using visible reflectance spectroscopy (VRS), there is currently no way to determine melanin distribution in the epidermis. We have developed a photoacoustic probe that uses a Q-switched, frequency-doubled Nd:YAG (neodymium, yttrium, aluminum, garnet) laser operating at 532 nm to generate acoustic pulses in skin in vivo. The probe contained a piezoelectric element that detected photoacoustic waves that were then analyzed for epidermal melanin content using a photoacoustic melanin index (PAMI). Melanin content was compared between results of photoacoustics and VRS. Spectra from human skin were fitted to a model based on diffusion theory that included parameters for epidermal thickness, melanin content, hair color and density, and dermal blood content. Ten human subjects with skin phototypes I-VI were tested using the photoacoustic probe and VRS. A plot of PAMI v. VRS showed a good linear fit with r2=0.85. Photoacoustic and VRS measurements are shown for a human subject with vitiligo, indicating that melanin was almost completely absent. We present preliminary modeling for photoacoustic probe design and analysis necessary for depth profiling of epidermal melanin.