Summary It is almost 50 years since antimetabolites were first found to have clinical antitumour activity, with Farber's discovery that aminopterin could cause remission in acute leukaemia. In the following 10 years, methotrexate, 6-mercaptopurine and 5-fluorouracil (5-FU) found their way into clinical practice. Subsequently, cytosine arabinoside was found to have activity in acute leukaemia, but, until recently, other significant developments have involved optimizing the efficacy of existing antimetabolites, including the use of leucovorin with methotrexate or 5-FU. Recently, new antimetabolites have become a fertile area for anti-cancer drug research. Gemcitabine (GEMZAR®) has emerged as an important new agent in several tumour types, including pancreatic, non-small-cell lung, bladder, breast and ovarian cancers. Capecitabine is an intriguing new prodrug, offering tumour selectivity and prolonged tumour exposure to 5-FU. More potent thymidylate synthase inhibitors have also emerged; raltitrexed is now commercially available for the treatment of colorectal cancer. Others under development include LY231514, which has other sites of action, hence the acronym MTA (multi-targeted antifolate). A novel target is glycinamide ribonucleotide formyltransferase (GARFT) and LY309887 and AG2034 are undergoing clinical investigation as GARFT inhibitors. A critical element with LY309887 appears to be co-administration of folate. It seems entirely possible that several novel antimetabolites will establish themselves in clinical practice in future for the treatment of solid tumours.