We have previously built an ultra-high resolution barcoding system that can be used to simultaneously track the evolutionary dynamics of millions of yeast lineages in real time. By evolving an initially isogenic barcoded population, we identified and characterized tens of thousands of mutations that provide a benefit to the initial genotype. Describing the mutation rate spectrum of these first beneficial mutations, the fitness effect of each mutation and the rate to that fitness effect, allowed us to predict the evolutionary dynamics of the population for ~80 generations. However, making predictions out to longer times is challenging because the evolutionary dynamics will depend heavily on second beneficial mutations and how these interact with first mutations. With this technology, second mutations are difficult to detect because they are most likely to occur in lineages that have already expanded to large sizes. Here, we develop a new double-barcode lineage tracking system that allows us to measure the fitness effects and occurrence times of thousands of first and second mutations in the same evolving yeast population. This system consists of a double landing-pad capable of accepting two large barcode libraries that can be inserted at different times during an evolution. The addition of second barcodes late in an evolution allows us to break up large lineages containing a first mutation in order to quantitatively study the second mutations that occur on top. Using this system, we are evolving presumably isogenic haploid and diploid yeast libraries containing ~200,000 first barcodes in carbon-limited media at two temperatures. At a time when most cells contain a first beneficial mutation but few contain a second, we will insert the second library of ~2M barcodes and evolve for another 200 generations. Amplicon sequencing of double barcodes will be used to track lineage trajectories following re-barcoding. Mathematical models will be used to estimate the occurrence times and fitness effects of second mutations, and how they depend on the first mutation. This high resolution and large scale evolution experiment will enable us to better understand macroscopic epistasis by systematically characterizing the relationship between first mutations and potential for further adaptation.