UK adolescents and young adults are facing increasing rates of mental health problems and extremely difficult economic circumstances. There is strong evidence that interventions to increase income during adolescence can mitigate conditions such as anxiety and depression. However, policymakers lack quantified risk differences in the probability of mental illness between different income groups by which to establish the prospective scale of mitigation. Here, we estimate risk differences using two longitudinal cohort studies: Millennium Cohort Study (cohort members born between 2000 and 2002) and Next Steps (born 1989–1990). We quantify the association between income and risk of depression at age 14, serious mental illness at age 17, and non-psychotic mental illness at age 25. We also conduct sensitivity analysis using numerous other markers of socioeconomic (SES) status. We estimate that those living in the poorest quintile group of households have a greater probability of mental health problems than the richest at ages 14, 17, and 25. We find that other markers of SES status—subjective financial strain, cohort member assessed wealth relative to friends, occupational class, and education—display steeper associations with mental health symptoms relative to associations between mental health symptoms and income. Our findings are likely to be conservative underestimates since they are likely to be attenuated by well-known and large measurement errors in both mental health problems and living standards during adolescence.