In some two-treatment clinical survival trials, a large imbalance in the allocation of patients to treatments will result in approximately the same power for the logrank test as equal allocation. This fact can be used in some trials to reduce significantly the number of patients allocated to a potentially inferior treatment.