Sixty-six compound including carcinogens, related compounds and non-carcinogens were dissolved in dimethyl sulphoxide or saline and injected intradermally into C57 Black male mice. Localised areas of depigmentation occurred with some chemicals . The effect was related to tumour-promotion with 56% of known tumour-promoters tested producing depigmentation compared with 24% of carcinogens and 17% of substances not known to be carcinogenic. Direct carcinogens ouch as alkylating agents which have both initiating and. tumour-promoting activity produced depigmentaion. Indirect carcinogens such as the polycyclic aromatic hydrocarbons and aromatic amines which require metabolic activation were not active. Some compounds known to cause occupational leukoderma were active. Amongst the tumour-promoters, tetradecanoyl phorbol acetate (TBA). carbon dioxide snow and croton oil produced the most prominent depigmentation. The dose producing the effect in half the injected sites (i.e. ED50) was 0 .018 μg for TPA, 0.18 μg for nitrogen mustard and 101 μg for chloroform. Other tumour-promoters causing slight or moderate depigmentation included Tween 20, saccharin and dodecane. The promoters anthralin and phenobarbitone did not produce depigmentation when injected in tolerated doses. Histological examination and electron microscopy of depigmented areas showed loss of pigment in affected hair and a decrease in melanocytes at the hair bulbs. the latent period for depigmentation was usually between three to six weeks, but ground Crocidolite fibres produced depigmentation after sixteen weeks. Other forms of asbestos fibres caused no depigmentation. The depigmentation of hair might be developed as a short-term screening test for tumour-promotion