It is increasingly apparent that target-based drug discovery is not delivering adequate numbers of clinical candidates for treatment of malignancies. The main problems encountered are target identification, validation and late stage attrition. New complementary techniques are required to increase the success rate of cancer drug discovery. Phenotype-guided discovery, using chemicai genetic screens, provides an alternative strategy that circumvents many of the issues associated with target-based discovery. Zebrafish is a vertebrate model uniquely suited to such high throughput screens, and establishing relevant neoplasia models would enable screening for novel anti-cancer therapeutics.