This paper, by constructing a CBRbased computational model, attempts to evaluate the validity of the Compensation Hypothesis, an SLA theory proposed by Professor Wang Chuming for the purpose of elucidating the cognitive mechanisms of Chinese learners of English. Instead of simulating the processes of phonological, lexical, semantic or syntactical acquisition, as is often done in first language acquisition research, the paper, based on the Compensation Hypothesis, examines the feasibility of modeling the cognitive mechanisms of bilingual contextual knowledge acquisition. This model holds that bilingual contextual knowledge can be best represented by the use of casebased reasoning, not rulebased reasoning. The architecture of the CBRbased model is detailed, related algorithm for the system proposed and a casebased corpus depicted. The computational model can account for some of the issues in SLA research, such as Chinglish, fossilization, etc. This methodology can also serve as an alternative research paradigm for SLA studies.