Adaptive evolution is a major driving force behind the phenotypic variation observed in nature, and recent advances in high-throughput sequencing now make possible the deep sampling of adaptive events during the course of a controlled evolution experiment. We developed a DNA barcoding approach to track over time the changes in population frequency of ~500,000 otherwise identical clones of the budding yeast Saccharomyces cerevisiae, and have monitored these clonal "fitness trajectories" for up to ~250 generations under a glucose limited regime (Levy, Blundell, et al., Nature, 2015). In this study, we have isolated 4,800 independently evolved clones from the 88 generation timepoint of the barcoded evolution, when most adaptive clones are likely to contain only a single adaptive mutation, and have developed an assay to assign fitness values to each individual clone in a highly parallel manner. This allows us to determine if the clone is adaptive (carries a presumptive adaptive mutation) or non-adaptive (carries either neutral or no mutations). We performed whole genome sequencing on hundreds of individual known adaptive clones, as well as many neutral clones as controls, and have identified many of the adaptive mutations that have independently arisen in these evolutions. There are two major classes of adaptive mutations: (1) self-diploidization, conferring an average fitness benefit of ~3%, and (2) mutations in the nutrient-responsive Ras/PKA and TOR/Sch9 pathways, conferring fitness benefits of ~5% to ~15%, and ~5% to 10%, respectively. Our large sample size and precision of measurement have allowed us to observe that the differential fitness benefits conferred by these mutations are dependent on the affected pathway, the individual gene, and even the type of mutation within a single gene. Additionally, we observe consistent differential fitness advantages for clones carrying mutations in paralogous genes; for example, IRA1 mutations confer on average a higher fitness advantage than do IRA2 mutations, while GPB2 mutations confer higher fitness benefits than GPB1 mutations. In summary, we have been able to link the specific molecular targets of adaptation to their fitness effects and build a comprehensive genotype-fitness map of the adaptive mutations that drove the initial evolutionary process in this system.