Skin tone reproduction has long been a challenge in image processing due to illumination by multiple sources in real‐world conditions. This paper describes an algorithm to achieve preferred skin tone reproduction. The work comprises two pivotal components including to develop: 1) a CCT-SPQ/D optimization model via controlled experiments to reveal the mapping relationships between correlated color temperature (CCT) and skin preference quality (SPQ) and chromatic adaptation degree (D), and 2) a novel white balance correction algorithm for skin regions under mixed illumination, which integrates local processing and spatial filtering with color temperature adaptive enhancement via the aforementioned model. Finally, a preference assessment experiment was conducted to demonstrate the superiority of the algorithm proposed.