Tattoo removal has long been a vexing problem. Although many methods have been involved, most of them are destructive and frequently cause scarring. However Q-switched ruby lasers have been successfully used to remove blue and black tattoos without the usual risks of textural change or scarring (Reid et al 1983). The major difference between this method and the others is that radiation from this laser induces preferential injury to cells containing tattoo pigment only. A major disadvantage of this method is the very high cost of the equipment, and in additional the red tattoos do not respond to red ruby light. This thesis investigates using a high power density xenon flashlamp for the removal of tattoos. The proposed method is based on the same principle as the Q-switched ruby laser, but has potential to remove various coloured tattoos, and to cost rather less than one tenth of the cost of a Q-switched ruby laser. In this thesis the spectral match between absorption of tattoo dyes and radiation of xenon flashlamp has been analysed. I suggest suitable treatment parameters for removal of tattoo using selective photothermolysis after calculation base on some histological studies. The theory of the xenon flashlamp system was analysed in order to design a flashlamp system, and some experimental trials on different pulse durations and brightness were carried out. I report on preliminary clinical trials on a volunteer's tattoos, using different pulse length and energy densities produced by various xenon flashlamps. The overall findings given by our preliminary experiments confirm that a xenon flashlamp with an appropriate energy density and pulse duration can selectively induce responses in a tattooed area by the mechanism of selective photothermolysis. These clinical trials suggest that an energy density of 8 J cm-² is probably the useful treatment threshold for 100 μs pulses.