Summary.-The effects of one large single dose of 5-fluorouracil (FU) have been compared to the same amount given in divided doses daily over a 3-or 5-day period.Comparison of the effects of single vs fractionated dosage was made on 2 types of experimental solid tumour with different growth, cell kinetic, histological and metastasizing properties. The tumour response was essentially the same for both the single and fractionated dose schedules.There were marked increases in animal mortality from drug toxicity following fractionated doses of FU compared to one large single dose. Mortality in animals with Tumour 3924A increased from 10% following one large single dose to 600/ for animals given daily fractionated doses for 3 days, and 80o/ for animals given daily fractionated doses for 5 days. Total marrow reserve was measured by the total DNA content of tibial marrow. The nadir of 6 days for loss of total tibial marrow DNA following one large dose of FU was increased to 9 days for both fractionated schedules of FU. The 3-day delay in recovery of the marrow prevented recovery within the time frame necessary for animal survival. The inference from these experimental cancertreatment studies is that daily fractions of chemotherapeutic agents such as FU result in increased morbidity and mortality, without benefit in the control of the solid tumour. The results question the advisability of the clinical practice of initially giving small daily "loading doses" of proliferation-dependent agents such as FU.These results emphasize the need for more precise information on the temporal relationship between the response and recovery of the host and the response and recovery of the solid tumour. They also emphasize the need for a better clinical understanding of the time sequence of solid-tumour recovery in relation to the time sequence of marrow recovery."ALTHOUGH 5-fluorouracil (FU) has been used clinically since 1957, there presently exists a wide divergence in opinion among individual investigators, institutions, and-cooperative groups about the optimal dosage regimen" (Ansfield et al., 1977). Initial "loading doses" of FU and other proliferation-dependent cancerchemotherapeutic agents are given daily or on alternate days and pushed to the point of mild toxicity. This approach to the clinical management of cancer patients uses, in most instances, host toxicity as the primary guide for the scheduling of the proliferation-dependent chemotherapy agents such as FU. Quantitative assessment of tumour response is often difficult or impossible. It is evident that more precise information on the temporal relationship between the response and recovery of the host and the response and recovery of the tumour is needed for a more rational design of clinical protocols.Studies on the effects of FU on experimental solid tumours have been carried out, to determnine quantitatively the effect of increasing doses of FU on tumour growth, animal morbidity and mortality. The ability to evaluate quantitatively the effects of different schedules of F...