Attention has been called to the distinction between "age incidence", which is a measure of risk, and "age distribution" which is not such a measure except under certain unusual circumstances which probably do not exist for any hospital experience in the United States. Examples to illustrate this distinction were drawn from death data for deaths attributed to leukemia and the lymphomas in the United States in 1949. The sex and race selection have been recorded for the types of leukemia and lymphoma separable in the sixth revision of The International List of Causes of Death. The age selection at death attributed to the numerically important of these causes has also been presented. To determine the age, sex, and race selection (incidence) of these diseases, with full confidence in adequacy of their classification, will require a cooperative study designed to apply uniform diagnostic technics to all cases occurring in some definable population such as a large city or a state. Data derived from individual hospitals or from literature summations are generally inadequate for this purpose.