Recently, Wheldon (1979) and Zeitz and McDonald (1978) have considered the effects of different fractionated irradiation schedules on a model tumour having properties similar to melanoma cells in culture. Using essentially similar assumptions, they came to different conclusions as to which of the four different schedules gave the lowest survival. Zeitz and McDonald (1978) considered the response of a totally oxic population of cells using the cell-survival data of Dewey (1971). They considered three forms of the equation for cell survival and showed that a multi-target model with initial slope gave the best fit to Dewey's data. However, regardless of the exact form of the cell-survival curve chosen, the greatest extent of cell kill was achieved when a few large radiation doses were used (3 × 12.00 Gy), and the least when a large number of small fractions were used (31 × 2.00 Gy). However, they demonstrated that when a large number of small radiation doses are used large differences in the estimates of cell kill can be obtained, depending on the exact form chosen for the cell survival equation. This is because survival is being determined by the shoulder region of the survival curve and that region is the least well described experimentally.