AbstractAppropriately defining and enumerating ‘fitness’ is fundamental to explaining and predicting evolutionary dynamics. Yet theoretical concepts of fitness are often hard to translate into quantities that can be quantified in wild populations experiencing complex environmental, demographic, genetic and selective variation. While the ‘fittest’ entities might be widely understood to be those that ultimately leave most descendants at some future time, such long-term legacies are hard to measure, impeding evaluation of how well more tractable short-term metrics of individual fitness directly predict longer-term outcomes. One opportunity for conceptual and empirical convergence stems from the principle of individual reproductive value (Vi), defined as the number of copies of each of an individual’s alleles that is expected to be present in future generations given the individual’s realised pedigree of descendants. Since Vi tightly predicts an individual’s longer-term genetic contribution, quantifying Vi provides a tractable route to quantifying what, to date, has been an abstract fitness concept. We used complete pedigree data from free-living song sparrows (Melospiza melodia) to demonstrate that individuals’ expected genetic contributions stabilise within an observed 20-year time period, allowing individual Vi to be evaluated. Considerable among-individual variation in Vi was evident in both sexes. However, standard short-term metrics of individual fitness, comprising lifespan, lifetime reproductive success and projected growth rate, typically explained less than half the variation. Given these results, we discuss what evolutionary inferences can and cannot be directly drawn from short-term versus longer-term fitness metrics observed on individuals, and highlight that analyses of pedigree structure may provide useful complementary insights into evolutionary processes and outcomes.