Survival rates were analyzed for an effectively population-based series of 77 children with acute lymphoblastic leukemia treated at one hospital during 1970-1981. Treatment was according to Medical Research Council protocols UKALL I to UKALL VII but with a great emphasis on compliance. The relapse-free survival rate was 54.5% at 5 years. The overall 5-year survival rate of 64% was substantially higher than the 47% recorded during a similar period over the rest of Britain. The difference between survival rates in this series and nationally was especially marked for children aged 2-9 years (76% versus 50%) and with white blood count under 10 x 10(9)/L (87% versus 57%). For both of these groups, the survival rates achieved were similar to those now being recorded for the UKALL VIII trial in which treatment is more sustained than in its predecessors and there is greater emphasis on doctor compliance with protocol. These results suggest that although the advantage of UKALL VIII over previous trials for poor prognosis patients may be attributed to the more sustained treatment, the improvement for good prognosis patients may be due to more rigorous compliance with protocol.