Skin pigmentation disorder is an abnormal melanin production condition that causes skin to appear lighter or darker. To assess disease severity, the concentrations of melanin pigment types namely eumelanin (dark brown-black pigment) and pheomelanin (red-yellow pigment) need to be determined objectively for assessing treatment efficacy of skin pigmentation disorders and effects of whitening cream. At present, the assessment of melanin types and is invasive; skin biopsy is conducted for chemical analysis of skin samples. Based on the Dichromatic Reflection Model, a system has been developed to measure the melanin pigments types from light reflectance of skin allowing non-invasive objective measurements. According to the Dichromatic Reflection Model, the skin reflectance (the diffuse reflectance) has been absorbed within skin layers before it reflects back to the surface. The skin reflectance is basically a function of scattering and absorption of skin optical parameters within various skin layers. The system consists of a spectrophotometer and a Graphic Processing Unit (GPU) workstation is developed. The spectrophotometer provides the skin spectral information (skin reflectance). The reflectance is then analysed using an inverse procedure; Monte Carlo simulation of light transport in multi-layered tissue or known simply as MCML, running on GPU workstation. An inverse procedure is applied to determine skin optical parameters (i.e., melanin types) by fitting reflectance measured by spectrophotometer with reflectance simulated by MCML. The system is validated with the Realistic Skin Model (RSM) of the Advanced Systems Analyses Program (ASAP) software. The system can estimate correctly with absolute error of 8.82 only. An observational study involving 110 participants having different skin tones based on Fitzpatrick classification conducted in Universiti Teknologi PETRONAS, Malaysia and University of Burgundy, France are found in line with those reported in the literatures indicating a correct estimation. Data parameters from a second observational study conducted at Hospital Kuala Lumpur and Hospital Serdang with 43 patients suffering from melasma (hyper-pigmented disorder) and vitiligo (hypo-pigmented disorder correspond to the spectral reflectance data of vitiligo and melasma). The developed system can accelerate execution of Monte Carlo for multi-layered skin tissues (MCML) to analyse skin pigmentation and determine melanin types to address the need for a non-invasive and objective evaluation of pigmented lesion and skin-whitening treatments.110 © 2018 by Taylor and Francis Group, LLC.