Academic underachievement refers to school performance which falls below expectations. Focusing on the pivotal first stage of education, we explored a quantitative measure of underachievement using genomically predicted achievement delta (GPA[delta]), which reflects the difference between observed and expected achievement predicted by genome-wide polygenic scores. We analyzed the relationship between GPA[delta] at age 7 and achievement trajectories from ages 7 to 16, using longitudinal data from 4175 participants in the Twins Early Development Study to assess empirically the extent to which students regress to their genomically predicted levels by age 16. We found that the achievement of underachievers and overachievers who deviated from their genomic predictions at age 7 regressed on average by one-third towards their genomically predicted levels. We also found that GPA[delta] at age 7 was as predictive of achievement trajectories as a traditional ability-based index of underachievement. Targeting GPA[delta] underachievers might prove cost-effective because such interventions seem more likely to succeed by going with the genetic flow rather than swimming upstream, helping GPA[delta] underachievers reach their genetic potential as predicted by their GPS. However, this is a hypothesis that needs to be tested in intervention research investigating whether GPA[delta] underachievers respond better to the intervention than other underachievers. We discuss the practicality of genomic indices in assessing underachievement.