This case study examines the shifting bilingual preference of three French/English bilingual children over a three-year period. It also clarifies the distinction between the many often misleading terms used to refer to bilingual preference (i.e., a bilingual's language choice). The children's fluctuating bilingual preference is accounted for in terms of three contextual domains: home, school, and community. The home domain was predominantly French-speaking, while the community domain shifted between predominantly English-speaking Louisiana and French-speaking Québec. The 10-year-old identical twin girls were in a French immersion program in Louisiana during the entire three-year period; their 12-year-old brother was not. A new, domain-sensitive longitudinal measure – the bilingual preference ratio (BPR) – was created and applied for each child using 36 months of weekly tape recordings of mealtime conversations. BPR fluctuations indicate that the greatest effect on the children's language preference was community immersion in the target language. However, the twins' markedly greater preference for speaking French at home in Louisiana is attributed to the influence of French immersion at school.