Since the late 1990s, thousands of fMRI studies have been conducted on different aspects of language processing in the human brain. The earlier studies were generally devoted to first language or monolingual processing, but the field has continued to expand to include both studies of a single first language, and bi/multilingual language processing in the brain. A modest number of fMRI longitudinal studies of second language acquisition began to emerge over the past 13 years. The following analysis uses the findings of these BOLD fMRI longitudinal studies of second language acquisition, including comparison with cross-sectional studies of L2 acquisition, to make recommendations for enhancing the research design and empirical measurements to facilitate new methodologies and approaches. Conclusions include a discussion of the utility of longitudinal studies, elucidation of the theoretical foundation of dynamic modeling underlying individual user variation in L1/L2 language processing, inclusion of a broader array of imaging techniques (structural DTI, resting state fMRI and functional connectivity), and the importance of proficiency measurements and proficiency testing as a part of research design.