Use of medical services by the elderly population is increasing. While rising numbers of elderly persons account for much of the overall increase, per capita increases in use of medical services within this group have been substantial. This phenomenon is commonly attributed to greater need due to the “aging” of the elderly population (i.e. an increased proportion of those 85 years of age and over) and to changing patterns of morbidity or to changing patterns of servicing the old and sick. Research to date suggests that the impact of the aging of the elderly population has been small. However, because the relationship between morbidity and rising utilization has not yet been studied, there is considerable debate about the extent to which increased “need” for care contributes to patterns of rising utilization. This research studied the impact of changes in numbers, demographics, and morbidity patterns of the elderly on per capita and aggregate consumption of ambulatory physician services at two points in time. Linked survey and physician claims data for representative samples of the elderly in 1971 and 1983 were used to study patterns of utilization of total ambulatory, consultative and non-consultative care by age and health status. The research found that per capita utilization of both consult and non-consult visits rose across all categories, with older individuals and those in poor health experiencing greater increases in utilization than younger and healthier individuals. At the aggregate level, a large percentage of the increase in utilization of medical services was related to increased numbers of elderly. Changes in ‘need’ attributable to aging of the elderly population and increased morbidity had only minor effects. At least 35 per cent of the increase in non-consult visits and 50 per cent of the increase in consult visits were not specifically related to changes in need of the elderly population and remain as unexplained changes in patterns of servicing individuals. While some of this increase was directed to individuals in poor health, the majority of increased servicing (32% of the increase in non-consultative visits and 58% of the increase in consultation visits) was directed to individuals in good health. Policy implications are discussed.