Methotrexate, an important agent in the treatment of childhood acute lymphoblastic leukaemia, has generally failed to induce dose-dependent cytotoxicity of patient-derived leukaemic blasts when tested in the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. This effect is apparently due to salvage from the medium, by surviving leukaemic cells, of metabolites such as hypoxanthine and thymidine. In an attempt to address this problem, we have examined the effect, on leukaemic cell populations, of enzymatically depleting these metabolites from the culture medium employed during the MTT assay, using xanthine oxidase and thymidine phosphorylase. Specifically we have assessed methotrexate cytotoxicity in the paediatric acute lymphoblastic T cell leukaemia, GKTL, which is maintained as a xenograft, and like primary leukaemias, has poor viability in vitro. Although little cytotoxicity of GKTL cells was observed when the MTT assay was performed in supplemented RPMI-1640 medium, dose-dependent cytotoxicity of these cells was clearly apparent when the same medium was enzymatically depleted. In contrast, the ID50 for methotrexate of control CCRF-CEM cells was unaltered in enzymatically depleted medium. In the absence of methotrexate, enzymatic depletion of the medium did not affect leukaemic cell survival. We are currently investigating the general applicability of this approach for assaying the response to methotrexate of primary leukaemia samples.