The relationship between the utilization of 3H-thymidine in situ ([3H]-TdR fractional incorporation or TFI) and tumour growth delay after treatment with various cytotoxic agents has been examined. It is shown that (a) it is not possible to predict tumour growth delay, or to select the most effective agent, from changes in TFI 1 day after treatment; (b) there is a good correlation between tumour growth delay and the time for recovery of TFI to the pretreatment level; (c) there is a relationship within a tumour line between the depression of TFI 4 days after treatment and growth dealy induced by the same treatment. This relationship appears to be independent of the mechanism by which the agent exerts its cytotoxic effect.