The present study examined sociophonetic variation in a small sample of Asian Americans in Boston, Massachusetts (N = 8; 4f, 4m; Mage = 23) representing four ethnic groups—Chinese, Filipino, Korean, and Vietnamese. Analyzing these speakers’ English production in tasks eliciting both casual and careful speech, we focused on four linguistic features comprising features observed in New England and in certain Asian American groups. Three features (L/R-CONFLATION, L-VOCALIZATION, and R-DELETION) were coded auditorily and one (LOW BACK RAISING of /ɑ/ to /ɔ/) acoustically; the dataset included approximately 1500 tokens of each feature, for a total of around 6000 tokens. Mixed-effects modeling results on the casual speech data indicated that Ethnicity was a significant predictor (p's < .05) of the occurrence of L-VOCALIZATION (M = 86.6% overall) and R-DELETION (M = 8.5%), but not of L/R-CONFLATION (which did not occur at all) or LOW BACK RAISING (which did not clearly occur in F1 or F2). Preliminary analysis of the careful speech data showed lower rates of L-VOCALIZATION (M = 70.0%) and R-DELETION (M = 1.5%) than in casual speech and, again, no occurrence of L/R-CONFLATION or LOW BACK RAISING. These findings reveal similarities and differences in speech production among ethnically diverse Asian Americans, point to a possible role of style, and highlight the need for further investigation of phonetic variation within this community.