A hypothesis that m-AMSA may have greater cytotoxicity in melanin-containing tumour tissues, because it may reversibly bind to melanin, leading to prolonged drug exposure, was examined. Clonal lines of B16 melanoma which differed widely in pigmentation level were selected by isolating artificial lung colonies and in vitro soft-agar colonies, and implanting them into mice. Excision cell-survival assays performed 24 h after drug administration showed that in vivo sensitivity to m-AMSA progressively increased as pigmentation level decreased, but that m-AMSA drug levels measured 24 h after treatment were much lower in amelanotic than in melanotic lines. In dose-survival studies the reduced sensitivity of melanotic cell lines was revealed as a large shoulder (Dq = 27 mg/kg) though the terminal slopes for melanotic and amelanotic cell lines were similar (D10 approximately 31 mg/kg). Time-course studies indicated that there was no significant loss of drug from a melanotic cell line for 72 h after drug administration, though in an amelanotic cell line drug levels fell 10-fold in 10 h. There was, however, no evidence for prolonged drug cytotoxicity in the melanotic cell line. Using a fractionated drug-treatment regime, the greater cytotoxicity of m-AMSA to amelanotic tumour tissue was confirmed in a non-invasive regrowth-delay assay.