Currently, most burn models for preclinical testing are on animals. For obvious ethical, anatomical, and physiological reasons, these models could be replaced with optimized ex vivo systems. The creation of a burn model on human skin using a pulsed dye laser could represent a relevant model for preclinical research. Six samples of excess human abdominal skin were obtained within one hour after surgery. Burn injuries were induced on small samples of cleaned skin using a pulsed dye laser on skin samples, at varying fluences, pulse numbers and illumination duration. In total, 70 burn injuries were performed on skin ex vivo before being histologically and dermato-pathologically analyzed. Irradiated burned skin samples were classified with a specified code representing burn degrees. Then, a selection of samples was inspected after 14 and 21 days to assess their capacity to heal spontaneously and re-epithelize. We determined the parameters of a pulsed dye laser inducing first, second, and third degree burns on human skin and with fixed parameters, especially superficial and deep second degree burns. After 21 days with the ex vivo model, neo-epidermis was formed. Our results showed that this simple, rapid, user-independent process creates reproducible and uniform burns of different, predictable degrees that are close to clinical reality. Human skin ex vivo models can be an alternative to and complete animal experimentation, particularly for preclinical large screening. This model could be used to foster the testing of new treatments on standardized degrees of burn injuries and thus improve therapeutic strategies.