Models of the evolution of life span predict, and gerontological studies show, a relation between nutrient use and life span. This study examines the role and comparative use of nutrients in long- and short-lived populations of D. melanogaster selectively bred for age-at-reproduction, without respect to metabolism or feeding rate. We test the hypotheses that selection for life span has favored the restriction of nutrient use and that the observed effect of low population density during development is a consequence of modifying nutrient use. The use of nutrients was measured here by the uptake of radiolabeled glucose, its incorporation into lipid and protein, and by the in vivo metabolic flux through the pentose shunt. Measurements show that uptake, incorporation, and flux are severely limited in long-lived stocks, compared to short-lived populations of the same stage and stage of development. Raising long-lived stocks at low population numbers relieves the restriction on metabolism in larvae, increasing incorporation and flux, and causes adult life span to decline. Larvae of long-lived populations appear to feed less actively, suggesting that the reduction in use of nutrients could simply be from reduced intake. Changes in total soluble protein correspond with measurements of uptake, incorporation, and flux. Soluble protein is substantially less in long-lived stocks, before pupation, but catches up in early adulthood to that in the short-lived populations. Despite different nutrient use by larval populations, only slight differences are found in development rate. They are insufficient to account for the differences observed in longevity.