This study investigated the processing of formulaic language (i.e., memorized chunks or conventional expressions) such as conversational speech formulas, idioms, proverbs, and pause fillers (e.g., uh, um) in three structured tasks and conversational speech, and examined how these fixed expressions differ across first (L1) and second language (L2) in bilinguals with Parkinson's Disease (PD). Previous studies reported an impoverished production of formulaic language, despite intact comprehension, in monolingual English speakers with PD. Additionally, it has been reported that bilingual persons with PD have greater impairments in their L1 than L2 compared to healthy controls regarding implicitly-learned grammatical knowledge. The current study aimed to examine whether formulaic language in L1 will also be impaired due to subcortical damage relating to PD. However, it was predicted that L2 performance will not be affected by PD, but only by L2 proficiency. Korean-English bilinguals with PD, who acquired Korean as L1 and English as L2, were recruited in this study. Two control groups comprised of healthy bilingual Korean-English individuals (matched controls) and healthy native English speakers were included for comparison. Their performance on four different tasks (comprehension, completion task, implicit/explicit knowledge on form, spontaneous production) was measured and compared across groups for analyses. Results demonstrated that bilingual speakers with PD had significantly impaired comprehension of formulaic language in L1 and had lower proportions of formulaic expressions in their L1 conversational speech compared with bilingual controls. Regarding L2, both bilingual participants with and without PD were comparable in their English performances across all tasks, but both groups performed significantly poorer in the structured tasks than native English speakers. Production of formulaic language in English (L2 for bilinguals) was similar across all three groups. These findings are in concordance with previous studies that support a proposal of L1 implicit language processing relating to basal ganglia, and further substantiate a subcortical role in the modulation of formulaic language. This study provides insight into a better understanding of the impact of PD on language functions and the neural basis of language learning.