Although numerous lasers with different wavelengths are available for laser hair removal, their use in individuals with dark-pigmented skin remains a challenge. The present study aims to develop a numerical heat diffusion model considering skin types over various wavelengths. This numerical mode uses Pennes approximation to represent heat from metabolism, blood perfusion and an external heating source. The heat diffusion model is experimentally validated by using agar-based skin tissue phantoms. Diode lasers with four different wavelengths were used with two antithetical skin models. The pulse width and beam spot size were set to 200 ms and 1 cm(2), respectively. Temperature distribution along the hair structure and skin tissue was examined to determine both thermal confinement and heat transfer to the hair follicle. Experimental results are well matched with the numerical results. The results show that for the light skin model, thermal confinement is well achieved over various wavelengths, and treatment efficacy is expected to be better at a shorter wavelength. Otherwise, for the dark skin model, thermal confinement is poorly achieved as the wavelength decreases (<808 nm) and the temperature gap between the hair tip and the hair root is significantly large compared with the light skin model, which may lead to adverse effects. We believe that the developed numerical model will help to establish optimal laser parameters for different individuals during laser hair removal.