In this study, we examine the largely neglected long-term effects of bilingualism for students with roots in immigration. Our central research question is whether students' bilingual proficiencies have an impact on their future earnings in the USA. For this purpose, we used two different data-sets, i.e. the National Education Longitudinal Study (NELS) and the Children of Immigrants Longitudinal Study (CILS). Based on Latent Class Analysis, we distinguish between three linguistic minority groups: limited bilinguals, balanced bilinguals and English-dominant group. For both CILS and NELS, the results of regression analyses show that balanced bilingual students earn significantly more as adults at the beginning of their career than those linguistic minorities who were dominantly proficient in English only. Even after controlling for cognitive ability, educational attainment and parental socio-economic status, the additional cost of complete linguistic assimilation is estimated at $2100-$3300 annually. The NELS-data also suggest that balanced bilingualism has an additional indirect effect through academic attainment. Policy implications of these results are discussed.